Quick BOI: A Reference Handbook for The Beginning of Infinity

A practical reference handbook for The Beginning of Infinity.

For understanding knowledge (epistemology):

What is it?
How do we get it?
What practical consequences does understanding it have?

Each new concept is a new section and new sections often build on old sections.

Graphs show those connections so one can see the supporting rationales quickly.

Foreword

David Deutsch introduces the concept that knowledge is information that tends to persist. He argues that a student creates new ideas during the process of learning, and that the information cannot be copied from the teacher. Ideas are therefore disseminated to people imperfectly, yet the good ideas will survive this error-prone process. For fun, this is a summary of my understanding from the Brett Hall’s excellent ToKast.

In this way, the ideas here are ones that have survived transmission, or been added by the student along the way. This work certainly contains Hall’s rich interpretations and remarks.

Karl Popper -> David Deutsch -> Brett Hall -> myself -> you

The Beginning of Infinity introduces a framework that feels actionable, clear and optimistic. The aim of this short guide is to:

Help me to understand and communicate the concepts better.
Help others do the same.

Each concept is reduced to basic elements and presented in three styles, with the content repeated in three forms:

Simple brief summary.

Plain language concept.

Short jargonised summary.

The content is based on Brett Hall’s reading of the Beginning of Infinity in his podcast ToKast. Chapters and content are therefore roughly aligned with those in The Beginning of Infinity.

Additional separate sections are:

A dissection of Moloch

The book source text and scripts for generating diagrams are located at https://github.com/ceramic-sf/quick-boi-source. All edits/PRs should be made in that repository.

Table of Contents:

QuickBOI - A Reference Handbook for The Beginning of Infinity

Chapter 1 - The Reach of Explanations

Realism

To believe that the world around us exists is both practical and rational.

We seem to exist in a real world where there are actions we can do to build and fix things. It is possible to embrace this idea deeply, and this has practical effects because it allows real progress in the real world and prevents becoming lost in esoteric discussions.

Realism is the acknowledgement objective reality.

Fallibilism

Everything may be wrong in some way. Good ideas are very useful, but they may always be replaced by better ideas.

Theories contain both truth and misconceptions. A theory may be good but later found to be improved by another, better, theory. So it is with every theory, and no matter how good a theory is, there is always room for another theory to be created that improves. Thus every theory can be thought of as having some true and false components.

Fallibilism is a foundation concept. People are subject to error, and so whenever we think that we have something true, we could be wrong about that. Theories contain both true and false elements. Newtonian physics was a great theory that was practical and contained some truth, it was replaced by Einstein’s General Relativity, which is more accurate. “By what means?” is the question to find how a theory may be fallible.

TODO

The problem with Empiricism

Empiricism ("seeing is believing") is wrong. It is not enough to see things to understand them, you must develop an idea about that thing and then challenge it.

Seeing something happen can be a powerful way to develop ideas about the world. That is not a good way to develop good theories. It is insufficient to use and trust observations alone. It is instead ideas and their critiques that lead to knowledge.

Empiricism (observations lead to knowledge) is refuted by fallibilism (any piece of knowledge may be wrong), for all observations are made by people in the end. Knowledge is formed first through conjecture then refutation.

connections_for_the_problem_with_empiricism

The problem with induction

You cannot make theories by looking at things and extrapolating. Firstly because you may not be seeing the whole picture, second because the future is different from the present.

Making extrapolations from observations is called inductivism. Like predicting black swans when you have only observed white swans. Repeated observations do not get you closer to something that you have not (or cannot) see. The issue is that you may only be observing a small part of a larger system.

Inductivism states that repeated observations allow us to create theories. It is refuted in two ways: The unseen represents the seen; The future does not look like the present.

The use of observations

We use observations to choose between theories that we have already guessed.

To observe or experiment is the mechanism to make a selection between theories we create. It is the process of comparing ideas about things to the way things actually are.

Observation does not lead to theories, theories are invented. Observations are used to choose the best theory.

The problem with justificationism

Don't believe theories that appear to be true. Use them, but look for better theories.

Justification is the idea that theories should be believed once they are supported by good evidence (justified). It is incorrect because theories will always contain some error. It is better to believe that theories may be “the best we have so far”, but they may be replaced by better theories later.

Justificationism is refuted by fallibilism.

connections_for_the_problem_with_justificationism

Truth

A truth is a statement of something that is real. A falsity is a statement of something that is not real. Things can have some true and false parts.

Truth is a regularity or law in nature. Something is a truth if it is the way something actually happens. Something is false if you describe how something happens, but that is not what really happens. A theory that predicts apples fall upward is false. A theory that predicts it falling is more true and a theory that does so more accurately is even more true. A theory with ‘pretty accurate’ predictions therefore has both true and false components.

Truth is a regularity in nature.

connections_for_truth

Incremental truth

We find theories with more and more ability to match reality (truth) over time.

Theories have both true and false false components because if a better theory is invented it has more truth than the original theory. The original theory still has some truth, even after the new theory is made. We work to incrementally increase the truth in theories, but you never know what hidden false part may be discovered because all theories are fallible.

Theories contain a mix of truth and falsity. We seek to make incremental increases in the truth in new theories.

connections_for_incremental_truth

Tradition of criticism

When people are encouraged to challenge old ideas, new ideas are made. If they are not, then old ideas persist, even if false.

The Enlightenment happened when society developed a culture of challenging the current theories. This encourages people to create new, better theories with more truth.

A tradition of criticism is the defining characteristic of the Enlightenment. A philosophical change leading to better theories.

connections_for_tradition_of_criticism

Testability allows falsification of theories

A theory is good if you can do something and check that the theory matches what you see.

Scientific theories are good if there is an experiment where some observation could render the theory false. Theories are not made by experience (this is inductivism), but experience is used to test theories. An example of a good theory is a caveman with a theory about a sharpened rock being better to cut than a blunt one.

Bohr had a theory that an element could turn into another element (“transmutation”). The theory was good because he could test the idea using observations (mass change after radioactive decay of radium).

connections_for_testability_allows_falsification_of_theories

There are two types of tests, crucial and non-crucial

A crucial test is one that allows you to choose between two competing theories. A non-crucial test points out some problems in the current theory.

The two types of tests (experiments) are:

Crucial test. An experiment that, if the results were “just so”, would falsify a theory. It allows you to choose between two competing theories. This is a crucial component of science. This is a cornerstone of falsification.
Non-crucial test. An experiment that if the results were “just so”, would merely make the theory problematic. However, with no better theory, it is still tentatively held as the best theory until a better one is invented. It is useful for directing research programs.

A crucial test allows a choice between two theories. A non-crucial test does not.

connections_for_there_are_two_types_of_tests_crucial_and_non_crucial

Good explanations

A good explanation is one that is hard to vary. It does not have multiple unrelated elements that can be modified.

Explanations are ideas about how things work. They need to be very narrow and precise if they are to be tested. An explanation that is easy to vary is hard to test. An example of a bad explanation is a caveman that predicts one rock will be better at cutting because of a complex theory involving where it was found, and what the weather was like at the time. While the two rocks may be tested, the result may be attributed to any number of the location and weather variables. A good explanation might involve the sharpness of the edge, and would be valuable in other situations, such as a wooden knife.

Good explanations ideas that are hard to vary while maintaining the ability to accurately describe how things work.

connections_for_good_explanations

Science is the search for good explanations

The purpose of science is to find good explanations.

The purpose of science is to find good explanations, not to predict things. A bad explanation can accurately predict things without representing the reality involved. Someone who watches many magic tricks might be able to predict what will happen in the trick. Predictions do not get at what is really happening on the stage, but you could find a good explanation that does and that is the goal of science.

Science is about finding good explanations. It is not about generating predictions.

connections_for_science_is_the_search_for_good_explanations

Good science

Good science is when an explanation is created and then tested. It is an explanation plus a testable theory.

Good science involves first developing an explanation about why something will happen, and then testing that explanation. This is only possible if there is a tradition of criticism. Doomsday prophets (world ends in 2 days) and gamblers (the next roll will be a 6) have testable theories, but this is bad science because it leads to explanations that are easy to vary (world actually ends in 4 days).

Good science is a good explanation plus a testable theory. The substance of scientific theory is explanation, and the explanation of errors constitutes most of the content of the design of any non-trivial scientific experiment.

connections_for_good_science

The problem with correlations

A search for correlations can result in findings that predict things well, but encourage thinking that does not match reality of how things work.

Search for correlations leads to testable theories without explanations, which is bad science. Correlations can be found between two variables without an explanation about why one thing causes another. An explanation allows the theory to be falsified.

For example consider that people living near power stations have worse health. If no explanation is created, a policy to ban living around power stations might be made. One theory is that radiation from the station affects health, you could falsify that theory by putting people near power stations and seeing if they had worse health. It turns out that the lower health was caused by properties being cheaper around power stations, and those people had less money for health care. So putting a wealthy people near the station would not cause poor health, thus falsifying the theory.

Correlations are easy to find and use but can discourage the search for and attainment of good explanations.

connections_for_the_problem_with_correlations

The problem with Bayesianism

Bayes' theorem cannot create new theories because it uses the past to predict the future.

Bayes’ theorem is a good statistical tool in narrow applications. It cannot be used to develop new theories because it requires looking into the past, which is a form of inductivism.

Bayesianism is refuted because it is a form of inductivism.

connections_for_the_problem_with_bayesianism

Occam’s Razor

The simplest option is the best.

The fewer components an explanation has, the more likely it is to be true.

The multiplication of components of an explanation beyond unnecessarily.

Knowledge

Something that can point itself at something and physically change it.

Knowledge is something that by itself can enact a real change at a destination. Knowledge has the unique ability to take aim at a distant target and radically transform it without having any impact on things in between. For example, cities are places where the landscape has been transformed as a result of knowledge. As long as there are people there to receive and use it.

Knowledge is something that if it points itself at a suitable destination, it has a transformative effect in reality at that destination.

Reach of explanations

Ideas have impact on the real world

Ideas “reach out” to solve problems beyond what they were originally understood or predicted to have. One explanation may result in dozens of explanations about seemingly unrelated phenomena. For example, the axis tilt theory of earth resulted in multiple explanations about things in the world.

Explanations have effects on multiple areas beyond the original context.

Chapter 2 - Closer to Reality

Fundamental theory

A theory that appears in many other theories.

A theory which lies beneath many other theories. An example is the axis tilt theory of earth, which appears in many explanations (weather, sunrise, star movements).

Fundamental theory appears inside many different theories.

Foundational theory

A theory that other theories depend upon.

TODO Check this. A foundational theory is one that required for subsequent theories to make sense. Fallibilism is foundational because it is the basis for many theories, but it is not a fundamental theory (a theory that keeps popping up everywhere).

Foundational is one that if it were removed, other theories would be affected.

connections_for_foundational_theory

The problem with seeking absolute truth

It is wrong to seek final answers.

To seek absolute truth is to be anti-fallibilist, because any final destination itself is fallible. Scientists should not seek final answers for this reason.

Absolute truth is refuted by fallibilist.

connections_for_the_problem_with_seeking_absolute_truth

The universe is our home

The universe is large, and the bigger the better, because there is more for us to use.

The whole universe contains resources for us. It is not there to help or harm us. That the universe is large is great news for our future.

The universe is a tool for us, neither friend or foe.

Instruments separate us to bring us closer

A telescope comes between you and the stars, but it brings the stars closer.

When you see the potential for life in a detailed photo of many stars, we feel closer than if we simply looked at the stars. Instruments between you and the thing being observed reduce the error your observations. This is because we have theories about how the instrument works. A computer telescope increases the separation even further, and reduces the error again. Humans look at the atoms of a computer display in a laboratory, but what they see is stars.

Instruments separate the observer, but because of our theory of how they work, they reduce errors in observation.

Humans can be creative even during boring tasks

Humans can have a sense of awe while doing some repetitive task, like counting stars. This leads to creativity, where an automated system will not.

Computer astronomers do not get a sense of awe like human astronomers. Computers play chess mindlessly, unlike humans. A human has the capacity to create new ideas even while doing mundane tasks.

Humans can produce creative outputs while doing automatable tasks.

Knowledge can be objective

Knowledge may or may not match reality, and there can be a real way test if it does.

Any knowledge may have some false part that can be identified and corrected. This is a real way that knowledge either matches or does not match reality. So knowledge can objective in this regard.

Knowledge is fallible, and being fallible implies error can exist. The discovery of the error can be objective, though some errors may be very difficult to discover.

connections_for_knowledge_can_be_objective

The problem with instrumentalism

If something gives good results, you may trust the machine not care how it really works. However if you have the wrong idea, you may never know what is really happening.

Instrumentalism is about being okay with some equipment that works in a mysterious way. Af the instrument works well (for example at making predictions), then that is useful, however it might reinforce an incorrect idea of how something works. It is better to discover how something works, otherwise you rely on empiricism.

Instrumentalism discourages the creation of good explanations. It is closely related to empiricism.

connections_for_the_problem_with_instrumentalism

Observation is theory laden

Before you make an observation, theories about what you will observe happen first. We know something about what we are setting out to do. Collecting evidence without being aware of what you might find will lead to bad data.

Creating new knowledge requires an increasing amount of theory in advance of our observations. A telescope with more components to correct for more errors first requires theories about those errors an how to correct them. Thus theory comes before observation, which is the reverse of observation-first empiricism.

Another example is to consider observing “visitors to the museum”. First you must develop theories about whether employees, side entrances and cafe-only visits are “visitors”. You have to have theories about what it is you are observing before you start observing.

Observations are laden with theories, which are used to make the instruments for observation. This refutes Empiricism which first requires observation.

connections_for_observation_is_theory_laden

Experimental evidence does not lead to new theory

New theories are made, then experiments are made to test them. New theories can come at any time or place by engaging creatively, in the experiment lab or at the beach.

Seeing evidence in an experiment doesn’t lead to new theories (that is inductivism). Theories are ideas created and then criticised. The experiment is designed around the theory, and the evidence that is collected is used to test if the theory should be rejected.

Any experience in life may lead you to encounter a problem, which you then may try to solve. Sometimes experiments throw up new problems that are unexpected. Creating a solution to that problem in the form of a new theory is not the result of the experiment, even though it happened around the experiment.

Experiments are a process to reject bad theories. Evidence does not generate theories. Performing an experiment may expose you to new problems to create solutions to (new theories), but this may happen in any situation and is not unique to experiments.

connections_for_experimental_evidence_does_not_lead_to_new_theory

Imagination

Science is about creativity and imagination first and criticism later.

Science (the search for good explanations) is about people using creativity and imagination to generate new conjectures. That is the first step, and later is followed by more rigorous and critical processes to throw away bad conjectures.

The starting point of science is the use of imagination to create new ideas. Explanations must be pondered and created before they can be rigorously examined and refuted.

connections_for_imagination

Error correction

Knowledge is about error correction, where incorrect ideas about the real world are identified and fixed.

Error correction is the process of working towards objective knowledge. With every error that is corrected, knowledge represents reality with increasing accuracy. This makes it more equipped to operate in the world, where systems that are able to navigate the real things in the universe are more effective and resilient.

For example, organisms that build better machinery are correcting the error that the inferior machinery had. This makes them more able to interact with the molecules they encounter.

(TODO Integrate constructor theory definition of knowledge here. Knowledge does not require an observer to operate with CT.)

Error correction is central to the creation of knowledge.

connections_for_error_correction

The problem with relativism

To make progress you must compare two things against objective reality. A claim that all ideas are equal prevents this type of real progress.

The idea that every truth is considered equal is a problem because it ignores the methods of finding knowledge, where ideas are criticised in order to refute them. If you cannot criticise an idea then you cannot make progress. So to consider everything as ‘truth’ and treat all ideas equally is to prevent the the means of error correction.

Relativism incorrectly considers ideas that match reality to be equal to those that do not.

connections_for_the_problem_with_relativism

Chapter 3 - The Spark

Humans are not the center of the universe

The universe is not built for humans, nor did they play a role in its formation.

Physically there is no center of the universe and humans have no special role in the story of the universe.

That we are not at a geometric center of the universe, or central to the formation of the universe is a refutation of anthropocentrism.

The problem with parochialism

Patterns around us should not be expected to apply to the whole universe.

Making the error that what is happening in one’s local environment happens to be of universal importance. It is easy to mistake the features of ones own environment (movement of the stars) as objective realities of the broader universe (the anthropocentric revolution of stars around Earth). Parochial etymology is from ‘parish’, as in confined to a small or narrow area.

Parochialism is when rules of thumb can be confused as universal laws.

connections_for_the_problem_with_parochialism

Significant (fundamental) phenomena

An idea is significant if common sense is not enough to explain it, or if it explains many things.

A theory is significant if there are no obvious day-to-day (parochial) things that explain it well, or if it is a fundamental (the basis for explaining many different things).

A theory is significant/fundamental if parochial theories are inadequate to explain it, or it appears in the explanation of many other phenomena.

connections_for_significant_fundamental_phenomena

Explanatory knowledge

Information about how something works, which can be used to transform the physical world.

TODO (make this better, see constructor theory). Explanatory knowledge is knowledge (something that can, when pointed, have a real impact) that is a good explanation (a theory about how something works, in a way that is hard to vary). So when the explanation is pointed at something, it causes that thing to be transformed.

Explanatory knowledge is an force-like thing that involves explanations causing physical effects in the world.

connections_for_explanatory_knowledge

People are special

People are more important to the universe than non-people entities. They can explain how things work and other things cannot.

People create explanatory knowledge. This makes them significant in the universe because of the impact they can have. While it is good to reject traditional anthropocentrism, humans are unique and more special than other non-people entities. So the Earth is a “hub” and is significant in the universe because people are here.

People are special in that they create explanatory knowledge. This makes humans cosmically significant.

connections_for_people_are_special

People can make environmental anomalies

Making extreme cold is an example of how special we are.

Humans can make temperatures well below what is known to occur anywhere else in the universe. This simple example shows how we are special in the universe. You could theoretically use this to find people (non-human people).

Creating extreme cold is only possible for people. A sign of uniqueness.

connections_for_people_can_make_environmental_anomalies

The problem with the principle of mediocrity

Society today looks at people as the cause of all the problems in the world, and that they are nothing special. Yet we are the only ones that can save Earth from a multitude of problems.

The principle of mediocrity is that humans are nothing special, and it is a commonly held truism (typical organism on a typical planet on a typical star). However because humans are people, we create explanatory knowledge, which is not a mediocre thing. Humans take our inhospitable planet and make it hospitable. In the future we will anticipate and prevent problems into the future.

The principle of mediocrity ignores that humans create explanatory knowledge.

connections_for_the_problem_with_the_principle_of_mediocrity

The problem with Spaceship Earth

Earth is dangerous not a safe spaceship. We survive despite the hazards here.

The Earth is not here to sustain us and to sustain life. While friendlier than other places, it is a harsh environment. Disease, fire, wind, cold, predators. We do the best that we can to protect us from the environment. 99.9% of all animals have gone extinct. “The Earth is the only thing that can sustain us” is an argument that ignores our capacity to invent technology that makes other places habitable and abundant with resources.

The Earth is incapable of supporting human life. It is not a benevolent spaceship that tends to our needs.

connections_for_the_problem_with_spaceship_earth

It is people that make Earth hospitable

People have the power to make unliveable places liveable.

While space would kill a human in seconds, Oxfordshire winter would kill you in hours. We have a biosphere that prevents this (houses), so we survive, but only because of our capacity to change the environment. Much of the environment would otherwise kill us, as it has killed nearly all of the organisms that have come before us.

The most hospitable parts of the biosphere on Earth are made possible by people.

connections_for_it_is_people_that_make_earth_hospitable

The environment is expected to change

The environment keeps changing.

The Earth has a changing environment, on all time scales. Evolution happens because of this change, as organisms respond to the changes (geological, weather, food, predators). So the environment will keep changing, forcing the current organisms to keep evolving. That is the natural state, one of change.

The environment changes continuously.

The Earth does not preserve species

The Earth does not stop species from disappearing.

The environment keeps changing and organisms become extinct. The ones alive today will also encounter a changing environment. The species alive now are well suited to Earth as it is today, but may be not in the future.

The Earth does not preserve species.

connections_for_the_earth_does_not_preserve_species

Wealth

Wealth is how much you could change if you were to try. It relates to the amount of things we are able to use.

Wealth is a measure of what can be transformed. It is a measure of what sort of things you could change or create in the world around you if you were to try. It relates to the quantity of resources we have access to, including technology. It depends on how much explanatory knowledge we have, for resources increase the more we know.

The repertoire of physical transformations that they would be capable of transforming. It is a quantity describing a combination of the physical resources we have and our knowledge to use them.

connections_for_wealth

The carrying capacity of the Earth for humans is unlimited

New inventions change what we use and need on Earth. It is possible that the Earth can accommodate us indefinitely.

Knowledge of artificial fertiliser enabled us to live at higher density. We cannot predict what further technology will increase the capacity of the Earth further. As our wealth increases, so does the carrying capacity of Earth.

There is no concrete limit to how many humans the Earth can support given the right knowledge.

connections_for_the_carrying_capacity_of_the_earth_for_humans_is_unlimited

The problem with an untroubled past

There is no point in the past that had better conditions for humans.

Some consider that at some point in the past, humans were living in a way that was superior to today. This ignores all of the advances that our technology gives us today. See the paleo diet for an example of this, or the Garden of Eden.

Romanticising the past ignores the problems that existed then, but which we have now solved.

The problem with middle world

Humans can understand anything that can be understood. We are not limited to a middle-range of concepts.

Some argue that humans are only capable of understanding a middle-range of phenomena. We evolved on Earth suited to perceive a small subset of all things (nothing quantum/microscopic, nothing galactic or at the speed-of-light). However we have demonstrated an ability to make progress in these domains, and we have a pathway to continue to comprehend more things.

Humans have the capacity to understand concepts outside the environment that we evolved in. We are not limited in capacity to understand.

We can transform the universe

We can change the universe if we want to, have enough wealth, and know how to do so.

There are only three things that limit our ability to transform the universe into a place filled with life:

Our desire to do so.
Our wealth.
Our knowledge of how to do so.

Our ability to make the universe better is determined by our desires, resources and knowledge.

connections_for_we_can_transform_the_universe

The universality of quantum computation

There is a proof what shows that the laws of the universe can be run on a computer.

Whatever the universal laws are, they must be computable. If computable, then a universal computer will be able to simulate those laws.

The universality of quantum computation was proven in a 1986 Proof by Deutsch. It showed that quantum computers are possible. The Church-Turing-Deutsch-Lovelace principle states that every physical process can be simulated.

Knowledge is created by conjecture and refutation

Knowledge is made by coming up with an idea, and challenging it to see if it is right.

Knowledge is created by creativity and criticism.

Knowledge is created by conjecture and refutation.

connections_for_knowledge_is_created_by_conjecture_and_refutation

Anything that is physically possible is achievable given the right knowledge

If something is possible, people can do it. Of all the possible things, there are none that cannot be achieved.

There is no thing that is technically possible but not practically achievable, given the right knowledge. This is because quantum computers can run anything and everything that laws allow). Any physical transformation is either:

Impossible because it is forbidden by forces of nature (faster than light travel).
Achievable given the right knowledge.

There is no third option, where there is some possible thing that people cannot eventually do. Hence it is possible for humans to do anything, like turn a galaxy into a giant city.

Any physical transformation is either impossible due to the laws of nature, or achievable, given the requisite knowledge.

connections_for_anything_that_is_physically_possible_is_achievable_given_the_right_knowledge

To code is to comprehend

A programmer understands what they code.

In order to write the algorithm for a computer program, the programmer has an understanding of what it is they are doing. So to write a program to do some complex task, is to comprehend that task.

Coding is comprehension.

Everything in the universe is comprehensible

Humans can understand everything because we can program quantum computers, which can technically run everything.

Quantum computers can simulate anything physically possible, and so if there is some hard concept, we can write a computer program for that thing and the computer can run it. So everything can be comprehended in this way. This means that technically, humans with a quantum computer can eventually understand everything physically possible because coding is comprehension.

Quantum computers have universal computation, and with them we can program any possible thing in the universe.

connections_for_everything_in_the_universe_is_comprehensible

People are entities that create explanatory knowledge

People are things that make explanations that can have large effects on the physical world.

People is a term specifically referring to entities that create explanatory knowledge. They are special because they can technically do anything that is physically possible. This includes humans and can include Artificial General Intelligence and aliens. While we evolved in a particular environment on Earth, our capacity to comprehend is not limited by that.

People are entities that create explanatory knowledge.

connections_for_people_are_entities_that_create_explanatory_knowledge

We know of three types of knowledge

Knowledge can come in the form of evolution, explanations or culture.

Genetic, explanatory and cultural knowledge are three examples of knowledge. (TODO Expand to constructor theory for a single definition here.).

Knowledge can be instantiated within genetic systems, explanations and cultures.

Humans can match a superintelligence

There cannot be a superintelligence that does things we cannot comprehend.

Being able to create explanatory knowledge is a power that enables us to understand anything. We can improve our performance by increasing the speed and memory of our computers, and with them we can eventually match the capacity of a superintelligence.

A superintelligence is merely a quantitative improvement over humans, not a qualitative one. We can eventually match any theoretical superintelligence, given the right knowledge.

connections_for_humans_can_match_a_superintelligence

There are no realms outside our comprehension

To say there are aliens we can never comprehend is the same as saying there is a god.

Given we can understand a superintelligent alien, there is no entity that exists outside our realm of comprehension. Any such thing is equivalent to a god, which is not a good explanation.

An alien we cannot comprehend is equivalent to there being a god.

connections_for_there_are_no_realms_outside_our_comprehension

Knowledge creation hubs have three requirements

An environment can support the creation of knowledge if it has physical things, energy, and away to represent evidence.

An environment can support an open ended stream of explanatory knowledge if it has:

Matter
Energy
Evidence (for testing theories)

People in this environment with these things could create more explanatory knowledge, and use that to survive, perhaps indefinitely.

An environment can support knowledge creation if is has matter, energy and evidence.

connections_for_knowledge_creation_hubs_have_three_requirements

Space colonies

Humans can go to thrive in many places in the universe.

Humans can go to any knowledge hub and start surviving there by creating more knowledge. This could be a redundancy in the case that something happens to Earth.

Humans can go anywhere that supports knowledge creation.

connections_for_space_colonies

Moral knowledge

Morality is about what we should want.

Moral knowledge is about what to want and what to strive for.

Morality is about understanding moral knowledge. What it is that should be sought.

The problem with suffering as ultimate morality

Making choices about what is the right thing to do needs to go beyond suffering, because we may one day solve all suffering.

If we were to eliminate all suffering in the future, we would still need to choose what to do next. This would involved making decisions without having the concept of suffering as a guide. Hence, some other system is required.

Morality should not be about suffering because in a post-suffering world there are still moral choices to be made about what to do next.

connections_for_the_problem_with_suffering_as_ultimate_morality

The problem with utopias

A future civilisation will always have some problem to be addressed.

A utopia has no problems. However, everything is fallible and so there will never be such a place. There will always be problems. This is a direction in which to more and provides a basis for morality. In any near-utopia, there will always be errors to correct.

Perfect utopias are refuted by fallibilism. Working on problems is a basis for morality.

connections_for_the_problem_with_utopias

Problems are inevitable, problems are soluble

We will always encounter problems, and they can be overcome.

We will always have problems, even if we approach a utopia like state. However, people create explanatory knowledge, they are capable of solving any problem. These together form a direction that we can always work towards. Finding new problems and then solving them.

Nothing is without error, but with the right knowledge, each error can be corrected. This can be a basis for morality.

connections_for_problems_are_inevitable_problems_are_soluble

Morality can be about solving problems

We will always have a direction to base our morality on if we simply choose to try and solve problems.

Unlike suffering, we will never solve all problems. This can form the basis of morality, because we should want to solve the problems we face, and we can. Solving problems will reduce suffering along the way (suffering being a problem we can choose to address).

Solving problems is a never ending quest of continual improvement. It is as concrete foundation for morality.

connections_for_morality_can_be_about_solving_problems

Knowledge creates accurate and improving representations

Knowledge can result in one thing being represented by another thing, with increasing accuracy.

The creation of knowledge exhibits a structure where two separate bodies (a quasar and a human brain) can both be a representation of the same thing, in different physical forms. Over time (through the creation of explanatory knowledge) the accuracy of the representation improves as errors are corrected.

An incremental increase in the accuracy of representations is a unique feature of knowledge creation.

connections_for_knowledge_creates_accurate_and_improving_representations

Chapter 4 - Creation

Knowledge tends to remain

Knowledge is information that causes itself to keep existing.

Knowledge tends to remain in the environment over time. Once it is embodied in a suitable physical form (biological organism, or human mind). This is because it contains truth that makes it successful in the real world.

Knowledge is information that is durable. The durability is a consequence of some degree of true representation of reality.

connections_for_knowledge_tends_to_remain

Knowledge is created in one way

Knowledge is made by proposing an idea and criticising it.

People and evolution both produce knowledge in the same basic process (first idea, second criticism), with variation-selection (evolution) or conjecture-refutation (people). Good variations/conjectures are hard to vary, which allows them to be selected/refuted through testing.

Knowledge is made by proposal and criticism in both evolution and by people.

connections_for_knowledge_is_created_in_one_way

Evolution has limited reach

The knowledge that evolution makes comes from random errors leading to accidental improvements. People can be more careful and can create ideas for a specific purpose.

Variations in evolution are random, but conjectures by the mind can be carefully crafted. Therefore, evolution has a lesser ability to affect the world (less reach).

Evolution creates knowledge blindly, which limits the ability for it to solve specific problems. People can intentionally focus on creating relevant ideas.

connections_for_evolution_has_limited_reach

Nothing is spontaneously generated

You cannot leave alone non-living stuff and expect it to become a complex organism.

Life cannot be spontaneously generated, because that implies the knowledge has come from nowhere. Life contains truths about the real world that are used to successfully interact with its environment. There is no known process for truths to be reliably spontaneously generated. Such a process would itself require knowledge about the real world (this pushes the same problem one level back). However ideas can be messily generated/created, with true and false components. Ideas can then be challenged to see if they match reality or not.

Complex organisms are generated through a process of proposal and criticism. Knowledge cannot come to be embedded in a complex organisms without this process.

connections_for_nothing_is_spontaneously_generated

A student is not a passive recipient of knowledge

To learn, a student must actively generate what they think a lesson is all about. If they guess well, the thoughts will survive subsequent criticism in various forms.

A student cannot passively absorb the learnings of a teacher. They must actively engage in the topic and generate their own conjectures during the learning process. To learn passively is akin to knowledge being spontaneously generated in their mind. The student generates an idea of the content, and tests that idea in various ways. This idea-creation is demonstrated when a new concept is learned without remembering a single full sentence from the teacher.

The bucket theory of the mind where students passively receive knowledge is incorrect. A student actively guesses the content and begins refuting that idea.

connections_for_a_student_is_not_a_passive_recipient_of_knowledge

The problem with designer theory

Things being "made by a designer" is not good explanation. Any designer must itself have a designer.

For a complex thing an explanation for its origin should not invoke a mysterious designer. The reason is that the explanation for how the designer came to be is then a problem. If a designer could “always exist”, then the complex thing might as well also have that explanation, removing the need for the designer in the story.

Explanations that invoke a designer fail to provide an explanation for the origin of the designer.

Systems can be designed at high levels, not low levels

A designer can be an okay explanation for something if there is a good explanation for how the designer arose from a knowledge-creating process.

A designer can be part of an explanation if it ends with a connection to systems that create knowledge. For example, “a person designed this watch” is okay because the persons mind used creativity and criticism to design the watch, the person themselves cam from variation and selection in evolution, which itself has no designer. An explanation is problematic if it ends with a designer, such as “a person designed this watch, evolution made people, but aliens designed evolution” is bad because now there is no explanation for how the aliens came to be. A resolution to this example might be an explanation for an evolution-like process that lead to aliens developing through variation and selection.

A designer is an okay term to describe emergent knowledge creators. However, both the action and the origin of the designer must be attributed to conjecture-refutation based processes.

connections_for_systems_can_be_designed_at_high_levels_not_low_levels

There are three options for origin of life

Life interacts with the real world, and therefore has embedded knowledge for doing so. The knowledge can come from making wild ideas and seeing how good ideas persist. Other explanations for the source of knowledge in life forms point to additional sources for knowledge.

A mouse contains knowledge embodied in its physical form. Each organ represents a sort of physical knowledge about the physical world. As do the proteins in that organ represent knowledge of real chemistry. There are three candidate sources for this knowledge:

Spontaneous generation (but a spontaneous system that reliably matches reality must have knowledge to be accounted for)
A designer (but designers need their designers to be accounted for recursively)
Knowledge creation through variation and selection (viable). Useful and useless things are both made. Only things that successfully interact with reality tend to persist over time.

Life contains implicit knowledge, obtained through variation and selection. Spontaneous instantiation of knowledge and designer-based explanations both recursively invoke additional knowledge systems that are unaccounted for.

connections_for_there_are_three_options_for_origin_of_life

Evolution creates knowledge

Evolution is based on errors that can be helpful or unhelpful. Helpful errors are useful and they tend to continue to exist.

Evolution is a process where nature makes errors in replication. Some errors result in systems that interact well with the real world and tend to survive. Hence, an error is like an idea for how something should work, and may be true or false. This is similar to the process that people use when they create explanatory knowledge where ideas are made and tested.

Random errors during biological replication are equivalent to ideas about what will work. Errors that tend to persist must contain some truth about the real world, and can be considered created knowledge.

The problem with Lamarckism

Animals look like they are evolving in a certain direction, but evolution has no goals.

Giraffes stretching to reach leaves results in longer necks (Lamarckism). The idea here is that evolution directs itself toward better or more complex things. However, evolution has no goal, and there exist many organisms who do not increase in complexity over time (e.g. simple bacteria).

Lamarckism is a form of inductivism, where knowledge (adaptations) is gained from experience (environment).

Evolution favours the genes that spread

Evolution favours the gene that spreads throughout the population.

A population of birds may end up containing a gene that reduces the survival of the species of bird as a whole. That individual genes are favoured (nest early to get best nest) means that eventually the entire flock may nest at a suboptimal time (before the food is abundant), perhaps going extinct.

Neo-Darwinism states that evolution favours genes not species. A population of replicators, subject to variation (such as through imperfect copying), will be overtaken over by those variants that are better than their rivals.

Evolution does not favour the fittest organisms

Evolution does not favour organisms with the 'best' abilities.

Evolution selects for the gene that spreads. Thus, the organisms that appear may not be the fastest, strongest, smartest or ‘best’. The organism is merely a vessel for the genes, and genes may help the organism to survive so that it (the gene) may spread through the population.

Evolution does not optimise organisms for the environment.

connections_for_evolution_does_not_favour_the_fittest_organisms

Learning and understanding is about imperfect copying

A student will understand a creative variation on what is being taught. They don't understand the exact same concept.

When trying to understand from a teacher, a student may understand it to mean what makes most sense to them, what they expect to hear, what they fear to hear, or some other interpretation. The expectation of what they are about to learn is an active conjecture by the student, not a passive process, and may differ from what the teacher intended.

A student always misinterprets the lessen. They learn by creating their own conjecture about the topic, and then criticising it.

connections_for_learning_and_understanding_is_about_imperfect_copying

Learning good explanations is more accurate than bad explanations

A lesson about a good explanation (precise theory) is learnt accurately or not learnt. A lesson about a bad explanation (changeable theory) is learnt either accurately or inaccurately.

A student is creating a lesson actively in their mind. If the content of the lesson is hard to modify because it is a good explanation, a very bad copy is not going to survive criticism in their mind. So students will learn good explanations, or realise they do not understand.

A student may learn a bad explanation, and because it is easy to vary, may change the lesson dramatically and have it still make sense under criticism. Hence bad explanations may evolve more rapidly from person to person.

Good explanations have the property that copying results in success or failure. Successful copies are very similar to the original. Bad explanations can be copied with large undetectable changes, and can become very different.

connections_for_learning_good_explanations_is_more_accurate_than_bad_explanations

Neo-Darwinism is falsifiable

Evolution may be falsified, either partially or completely.

An observation that an organism is only favouring the optimal organism, rather than selecting genes, would be a refutation of evolution by natural selection. There are also different aspects of the theory that could be falsified, without refuting the whole.

Evolution may be falsified.

connections_for_neo_darwinism_is_falsifiable

The problem with the fine tuning of physical constants

The universe has delicate parameters we cannot explain. However not understanding does not imply there is a designer or god.

There are about 30 physical constants that govern the rules of the universe. It is predicted that if they were slightly different values, then life could not form (e.g., planets would not form etc.). Thus it appears that they have been tuned for some purpose (a designer has hand tuned our universe). However, because we do not have a good explanation, does not imply that such an entity exists. We encounter problems and problems are soluble. So while we do not currently know the explanation, someday we will.

The absence of a good explanation for the physical constants does not imply the existence of a designer.

connections_for_the_problem_with_the_fine_tuning_of_physical_constants

Knowledge is created and is unpredictable

Scientific discoveries are created by the scientist. They cannot be predicted, otherwise the predictor would then be the discoverer.

The creation of knowledge uses creativity and there is no way to predict that. If there were, then the process for predicting the content or consequences of that discovery would itself be that discovery. While the universe is deterministic, there is no known means to predict this emergent creativity. Hence knowledge creation is unpredictable.

Knowledge is created in an unpredictable way.

connections_for_knowledge_is_created_and_is_unpredictable

Knowledge creation cannot be predicted

We cannot predict what knowledge will be created.

The most important of all limitations on knowledge creation is that we cannot prophesy. We cannot predict the content of ideas yet to be created, or their effects.

Knowledge prediction is not possible, it is a form of prophecy.

connections_for_knowledge_creation_cannot_be_predicted

People have free will

People create knowledge in unpredictable ways, which means their thoughts are not predictable. They have free will.

knowledge is unpredictable, even though it is happening in physical objects governed by deterministic laws of the universe. So people can create things that are inherently unpredictable, and this is the basis for free will. Free will is due to the existence of emergent levels of creation in the mind. So while the mind is made of deterministic particles, the ensemble acts in a way that is not deterministic as a whole.

Free will is a consequence of the unpredictable nature of knowledge creation.

connections_for_people_have_free_will

Chapter 5 - The Reality of Abstractions

What is real?

Real things are things that we use in our explanations for how things work.

For our best explanations, anything that they refer to must be real. For example, the explanation for matter requires that we consider electrons to be real (matter -> atoms -> electrons).

If an entity is referred to by our best explanation in the relevant field, we must regard it as really existing.

connections_for_what_is_real

Abstractions are not physical

Things can be physical or abstract (non-physical).

Material, physical things are not enough to capture all things. Abstractions are ideas that not physical. They are patterns within those physical things.

The way in which physical things are arranged is not haphazard. Abstractions can cause things to happen.

Abstractions are real

Abstractions are real despite not having a physical form.

The next prime number exists, but it is not written or known anywhere. It can cause it can cause people to go out and find it. So it is an abstract idea, not physical but real.

The reality of abstractions is demonstrated by the observable effects they can have on physical things.

connections_for_abstractions_are_real

Emergence

Complex systems can produce simple systems that are separate from the underlying parts.

We cannot simulate fully all the particles in a boiling kettle, yet we can accurately understand the behaviour of the system as a whole (we know when the water boils). Free will may be a consequence of this wider view of a complex system. There are water and metal particles, yet the idea of a boiling kettle is abstract.

TODO

connections_for_emergence

The problem with determinism

A best explanation is sometimes the obvious high-level explanation, like why a statue exists. You would explain it in terms of society or culture, not atoms.

You can explain why something happens in terms of its particles, but there is sometimes a better explanation. Thus explanations of the emergent nature of a system (the story of a copper atom in the statue’s nose) include its context and are better than the particle-level explanation.

TODO

connections_for_the_problem_with_determinism

Dualism involves physical and abstract coexisting

Abstract concepts are real, and coexist with the physical. Hence the mind can have physical and non-physical parts.

A computer is built out of dominoes that computes if 641 is a prime number. Why does the output domino never fall results in two answers:

The preceding domino never fell
641 is prime

The former is true but is missing something. The second is the more useful answer, but does not reference anything physical, merely the abstract concept of primality (not part of the laws of the universe). Hence, the abstract and physical coexist (dualism).

Abstract concepts and physical objects coexist, and is an argument for dualism.

connections_for_dualism_involves_physical_and_abstract_coexisting

Knowledge is based on emergent phenomena

Knowledge is based on the simple patterns that emerge from complex systems. The explanation for water boiling is not about individual molecules.

All knowledge consists of and refers to emergent phenomena. The details of the knowledge are about high-level concepts, not lower-level deterministic concepts, as appropriate for the topic. The idea of water boiling does not refer to individual atoms or subatomic particles. So the particles and the idea both separately exist.

Knowledge is based on and consists of emergent phenomena.

connections_for_knowledge_is_based_on_emergent_phenomena

Theories replace each other, yet truth growths

Better ideas don't sit atop old ideas. They push the old to the side. Though the old likely contains some useful portion.

Each new theory that replaces an older one does not build upon it. It casts the old aside. Gravity went from being not-force-based to force-based to not-force-based (Kepler, to Newton to Einstein). Yet each old theory still holds some portion of useful truth.

New theories replace old ones, yet the old contains some truth.

connections_for_theories_replace_each_other_yet_truth_growths

There is no absolute truth

Things can seem true, but there is always some error that can be improved. So a better description is "our best effort at truth".

Our best theories, while appearing true, are flawed in yet-unseen ways. Old theories can still produce “pretty good” predictions, and still contain some truth. So our best theories are not true, but our most true theories.

We can only have provisional truth. Everything is fallible.

connections_for_there_is_no_absolute_truth

Knowledge, while abstract, causes physical effects to replicate

Knowledge is not physical (an idea for how stars work), but interacts with physical things (a brain).

Knowledge units are abstract replicators. They use the real world to achieve their replication. Something abstract (non-physical) such as knowledge or genes are causing something to happen (physical). Knowledge can use (and hence affect) our brains to get themselves replicated.

TODO Mix with constructor theory.

Pieces of knowledge are abstract replicators. An abstraction that causes physical change.

connections_for_knowledge_while_abstract_causes_physical_effects_to_replicate

You can decide what should be done

It is possible to know what should be done in life. When knowledge doesn't match reality, that is a definite direction to take action toward. One ought to solve real problems.

The phrase “you can’t derive an ought from an is” is a statement that you cannot know what should be done based on what exists. This usually leads to either dogmatism (following rules without criticism) or Relativism (only take actions that are good relative to something else). However problems really exist, and it is reasonable to state that they ought to be solved. This is a concrete objective framework for morality.

That problems really exist is the basis for determining what actions should be taken. There is an objective morality based on solving problems. This removes the need for moral dogmatism and relativism.

connections_for_you_can_decide_what_should_be_done

The problem with morality of effective altruism

A stronger framework for decision making is to solve problems generally, rather than trying to help people. Working to fix errors in our knowledge is a more robust and general approach that will also help people along the way.

Deciding what to do based on what maximises other peoples preferences ignores that there is objective truth, and that there are always problems/errors to solve. This can be the basis for objective morality. To target preferences is to avoid a more fundamental approach because suffering is a problem in a bigger group of problems.

It does not make sense to determine right versus wrong using their utility in meeting people’s preferences. Objective problems exist, where knowledge does not match reality.

connections_for_the_problem_with_morality_of_effective_altruism

Brains are universal turing machines

The brain can compute anything given enough time. A demonstration of this is writing basic computing elements on a piece of paper.

A turing machine is a computer that can compute anything, given enough time. The core ingredients are read, write and erase. Brains can do these with a pen and paper, and are hence turing machines. This is in addition to their ability to create explanatory knowledge.

The brain is a universal turing machine because it has the ability to read, write and erase symbols on paper.

connections_for_brains_are_universal_turing_machines

Proofs are computer programs

Mathematical proofs involve following steps to reach a conclusion. A computer program can do this.

Proofs are systems that involve following steps that produce conclusions. While often expressed as thoughts or symbols on paper, they can be expressed as computer programs.

Proofs are equivalent to computer programs in that they involve inferring conclusions from premises by following logical steps.

Mathematical proofs are subject to error

While math may be pure, we can only access it with error-prone physical things (computers and brains).

Mathematical proofs are computer programs that operate on physical objects, such as computers or brains. These objects are subject to error by nature of imperfect composition and external influence. Hence, proofs are subject to error.

Mathematical proofs are limited by the physical systems in which they are instantiated.

connections_for_mathematical_proofs_are_subject_to_error

The problem with getting purity from mathematics

Mathematics is the study of absolutely necessary truths. Necessary truth is merely the subject matter of mathematics, not the reward we get for doing mathematics.

Mathematical objects are things that are accessed via proofs, which are subject to error. So while mathematics may be pure, interacting with mathematics is an error-prone process. For example, creating a mathematical theorem requires a person to reason about the topic. This reasoning is a process that happens in the mind and is subject to error.

That mathematics can provide absolute truth (platonic idealism, or “the mathematicians misconception”) is refuted by understanding that mathematical proofs are physically instantiated fallible programs.

connections_for_the_problem_with_getting_purity_from_mathematics

The problem with mathematics being physical

Mathematics has physical and non-physical (abstract) parts.

Mathematics has abstract components, which are not instantiated in physical objects, yet are real.

That mathematics must exist in physical forms (physicalism or materialism) is refuted by there being abstract components. Both are a form of rationalism or empiricism.

connections_for_the_problem_with_mathematics_being_physical

Morality has a reality independent of the laws of physics

Morality is about making progress. This involves creating that have more truth in them. This could apply even in another universe with wildly different laws.

If a morally right action is to solve problems, then morality is tied to finding and fixing errors. This ties morality to the concept of truth, where to solve a problem is to make new knowledge that better reflects reality (a solution is creation of knowledge that contains more truth about the real universe that previous knowledge).

So even if there was another universe with different reality (different physical laws), the same process could be applied. This suggests that this definition of morality could be independent of the laws of physics.

Morality based on truth discovery (though problem solving) is a robust definition, in that it could apply in universes with different laws.

connections_for_morality_has_a_reality_independent_of_the_laws_of_physics

Chapter 6 - The Jump to Universality

Universality is about doing anything and everything

Systems that can be used to represent anything are universal.

If something is universal, it can do anything and everything. Alphabet-based languages can be universal, meaning they can be used to represent any concept. A pictographic language (one picture per word) is not. Other universal systems associated with the enlightenment include moveable-type, justice, legitimacy, morality, matter and motion.

A system is universal if it contains enough expressibility to represent anything.

Computers are universal

Computers can represent any physical thing, even physical things like biology.

A computer is universal because it has basic components that makes it able to run any program. They could even simulate an organism and its environment. However, some programs could be infeasibly slow.

Computers have the property of universality.

connections_for_computers_are_universal

Universality allows room for creativity

Being able to use a system for any possible task means it will be useable even after new ideas are created.

A universal system allows room for unpredictable things to be described with it in the future. The creation of new knowledge cannot be predicted, and so a universal system retains the ability to be useful in the future.

Universal systems retain the capacity to represent knowledge that is not yet created.

connections_for_universality_allows_room_for_creativity

AGI is about creativity

An artificial general intelligence has creativity that can be used for any task it encounters.

An artificial general intelligence (AGI) will have the capacity to solve problems that it was not programmed for. This is the property of universality of problem solving. An AGI is therefore not predictable, because it would create knowledge. This separates it from other sophisticated, but narrow, AI programs.

AGI is defined by its creativity. It can solve problems not described in its code.

connections_for_agi_is_about_creativity

Computers can simulate a human brain

If it exists, a computer can simulate it, including a human mind.

Computers are universal simulators, they can represent other things abstractly. Like how a person can have an abstract representation of a quasar in their mind, a computer can too. This means that a computer can have a representation of a human mind, and thus simulate it like it would a quasar.

Universality of computation means that anything physical can be simulated, including the human mind.

connections_for_computers_can_simulate_a_human_brain

Computers can be conscious

A computer that simulates a human mind can be conscious.

Given that a computer can simulate a human mind, computers can be conscious.

Computers can be conscious, because they can simulate human brains, and we are conscious.

connections_for_computers_can_be_conscious

Quantum computers are possible

Computation is a branch of physics, so anything that is physical can be computed. A quantum computer can be built to do this.

Quantum laws of physics can be used to show that computation is a branch of physics. Computers are made of atoms, which obey quantum theory, so there is a quantum theory of computation. This showed that computers are universal simulators and that quantum computers are possible to build.

The proof in 1985 by Deutsch of the universality of quantum computation is evidence that a computer composed of quantum elements can perform quantum computation.

connections_for_quantum_computers_are_possible

Error correction is required in computers

Computers need to check that the parts they are running on do not malfunction and alter the values being processed.

Computers have hardware with imperfect components, outside influences and thermal fluctuations that disrupt the computation. They must be checked for and corrected. A universal computer is required to run lengthy and precise algorithms so this is very important.

Universal turing machines require error correction.

connections_for_error_correction_is_required_in_computers

All universal computers are digital computers

There are no universal computers that are built from parts that have ranges of values. They must all have definite on-off states.

There is no such thing as a universal analogue computer. Error correction is required to run lengthy and precise algorithms. In an analogue computer the errors expand and cause the program to malfunction. Thus the parts that do the computation must not be a sliding scale of continuous values (0.3, 0.34, 0.345…), but instead have well defined values (such as 0 or 1).

Universal turing machines cannot have continuous computational elements, they must be discrete (such as binary).

connections_for_all_universal_computers_are_digital_computers

Human brain is a digital system

The brain can compute anything, so it must have error correction capacity to run those computations. Analogue systems do not allow this, and so the brain must be digital.

The human brain is a universal computer, and as all universal computers are digital, the brain is a digital computer. There must be binary on-off states that allow error correction, and indeed neurons behave in an all-or-nothing way, where they are either on or off.

The brain must be a digital computer because it has the property of universality, which requires error correction that analogue computers lack.

connections_for_human_brain_is_a_digital_system

Epistemology is digital

If your system for growing knowledge allows you to reject bad ideas, then your knowledge can grow forever (like a computer that fixes its errors to run long programs). If you keep bad ideas, they may build up and interfere the growth of knowledge (like a computer that never fixes errors and can't run long programs).

Epistemology the theory of knowledge. Knowledge creation is about error correction where new knowledge allows for old knowledge to be discarded. This is possible because part of the old knowledge is shown to be false. This system of keeping or discarding is binary, and thus digital. If epistemology is digital, it must run on a digital system, such as the human mind.

For comparison, other analogue systems for epistemology, instead retain both new and old theories rating them on degrees of truth, or probabilistic truth. These fail to have a method to correct errors and are therefore bounded rather than universal.

Popperian epistemology is digital, meaning that it can correct errors. Other kinds of epistemology are analogue and their scope is bounded not universal.

connections_for_epistemology_is_digital

DNA has universality

DNA is a system that seems to be able to specify any sort of organism on earth.

DNA is a system that codes all organisms on Earth, and appears to have the ability to code for any type of organism. In this way it is universal for life of the sort we have on Earth (proteins not silicon based life).

DNA is a universal code for life.

People have universal explanatory power

People can create explanations for anything. They are universal explainers.

People create explanations by having ideas. Brains are universal, which allows humans to create any sort of idea, and hence any explanation.

People can explain everything that is explicable.

connections_for_people_have_universal_explanatory_power

Constructors

TODO

TODO Constructor definition here.

TODO.

Universal constructors

TODO

TODO Universal constructors definition here.

TODO.

People are universal constructors

TODO

By extension of being universal explainers, they are also universal constructors.

TODO.

connections_for_people_are_universal_constructors

Chapter 7 - Artificial Creativity

A simulated mind is a mind

If you run a program that does what a mind can do, it is itself a mind.

A simulation of a mind is a mind. This is because a mind is an abstract object. However, a simulation of a physical thing (a bullet), is not that thing.

A simulation of an abstract thing is itself that abstract thing.

Qualia

The experiences of some thing, like the sensation of seeing the colour blue.

Qualia are the experiences that people seem to have. We seem to have them, but cannot explain them.

Qualia are a hard to characterise phenomenon we seem to have.

Turing test

The turing test states that if a program behaves like a human that is enough to call it intelligent.

A human judge tries to determine if a text-based program is being controlled by a human or not. If the judge cannot tell, then the program wins the prize.

A turing test is a proxy test for an artificial general intelligence.

If you can’t code it you don’t understand it

If you can't write a program for some thing, then you do not understand it.

If there is a concept one understands, that concept can be written down as a program. If the program works as expected, then the concept was understood. For example, we can simulate flying machines and therefore understand them. If we cannot yet code consciousness, we do not yet understand it.

To code something is to understand it.

We seek a good explanation of intelligence, not merely a winner of the Turing test

The Turing test helps us find AI, but it is not our end goal. We want to see how it works.

The purpose of the Turing test should be to understand how a winning program ended up winning. This would involve seeing the program, to discover a good explanation of what is happening. Otherwise there could be some sophisticated hoax. In this way, the Turing test is a way of weeding out hoaxes. There is no real need for the test, you can just publish the algorithm for the AI and let people see and test it.

The goal of the search for AI is to find a good explanation for how AI works.

connections_for_we_seek_a_good_explanation_of_intelligence_not_merely_a_winner_of_the_turing_test

An AGI should be able to explain itself

An AGI should be able to explain how its own program functions.

Given that an AGI has the capacity to create any sort of knowledge, it would be able to create an explanation for how its own program works.

AGIs should be capable of creating an explanation for the mechanism of its own program.

connections_for_an_agi_should_be_able_to_explain_itself

AI is categorically different to AGI

AI we have today is good at some things, but apply itself to any problem.

Programs that do narrow programs very well will always have domains they are incapable of solving. This is because they do not have the property of universality. In contrast, an AGI is defined by its ability to create new explanatory knowledge and has the capacity to solve problems in any domain.

AI has special purpose thinking, AGI is general purpose thinking including creating explanations.

connections_for_ai_is_categorically_different_to_agi

An AGI is likely to be consciousness

It seems unlikely that you could have consciousness without being able to create explanations about how things work. They seem to go together, at least in humans. Hence they likely are going to happen together in computers.

Humans have consciousness and the capacity to create explanatory knowledge. No other entity has either of these things, and while we do not understand consciousness, it is reasonable that they are connected. If connected, coding an AGI would make it conscious, rather than there being category between AI and AGI where programs could have consciousness but not be able to create explanatory knowledge.

The jump to universality for a universal explainer constructor appears to be associated with consciousness.

connections_for_an_agi_is_likely_to_be_consciousness

A sense of free will may be free will

It is possible to feel like you have free will, and that may be correct. If meditating creates a sensation that you have no free will, this is not rule out free will existing.

Meditation can separate the thoughts you have, and this can give a sensation of a lack of free will. Focusing on nothing can produce that sensation of no freedom to choose. However when engaged flow states, we can also get a sensation that we do have free will in our actions. Given that both sensations can be produced, it is reasonable to think that people have free will..

Meditating on thoughts can give a sensation of no free will, but that may not be evidence for a lack of free will.

connections_for_a_sense_of_free_will_may_be_free_will

Creativity gives people their ability to create explanatory knowledge

Creativity allows us to come up with new ideas to test. Some ideas will end up surviving the tests to become knowledge.

The ability of humans to generate new creative concepts is central in our creation of explanatory knowledge. It is creativity that is critical to the conjecture phase of knowledge creation.

Creativity is the production of ideas and conjectures without refutation. It is critical for the creation of knowledge.

connections_for_creativity_gives_people_their_ability_to_create_explanatory_knowledge

Genes do not encode specific mental capabilities

It is interest in topics, rather than genetics, that determines ability in different disciplines (mathematics, history or language)

The brain is universal and can be directed to think about any discipline. This includes those that are not yet invented. It is extremely unlikely that there are genes for specific domains such as mathematics or language.

The mind is universal, and so domain specific genes are unlikely because they are redundant.

connections_for_genes_do_not_encode_specific_mental_capabilities

Biological systems are a type of knowledge

Evolution by natural selection involves the DNA and the hardware that the DNA runs on.

Biology consists of physical forms in addition to the DNA that encodes specific genes. Both have evolved and represent truths about the world. For example the apparatuses in each cell have good models for how molecules behave and interact. This knowledge is implicit in that it is embedded in the system.

The knowledge produced by evolution is embedded implicitly in the entire organism.

connections_for_biological_systems_are_a_type_of_knowledge

We do not know how geochemistry becomes biochemistry

We don't know how initial compounds first started forming the basic ingredients for life.

We do not know the steps that led to chemicals start organising and accruing implicit knowledge. We have yet to discover how the first basic mechanism of variation and selection worked.

The origins of biochemistry are not known.

connections_for_we_do_not_know_how_geochemistry_becomes_biochemistry

Life may be easy to make

After a harsh asteroid period, we find that life quickly forms. So life may be easy to form in general in the universe.

Life seems to appear straight after the hot unfavourable asteroid (late heavy bombardment) period was over. This may be evidence that life tends to form readily on Earth-like planets.

Life started very quickly after the Late Heavy Bombardment.

connections_for_life_may_be_easy_to_make

The problem with evolutionary algorithms

Current evolutionary algorithms are not true evolution. The code it is written in introduces knowledge indirectly from the programmer.

Existing algorithms work toward some future that the programmer has implicitly directed the evolution towards. This implicit knowledge has reach which leads to the success of the algorithm. In fact all that knowledge produced by traditional evolutionary algorithms, was created earlier and elsewhere. The code contains the solution, even if the programmer cannot think of the specific answer.

True evolutionary algorithms require that the program has no goal. In all other goal-based algorithms the code has implicit knowledge with reach that has come from the programmers mind.

connections_for_the_problem_with_evolutionary_algorithms

We do not know how to program evolution or the minds of people

We know that computers can run knowledge-producing programs (for evolution or people's minds), but we do not know how to write the programs yet.

While evolution and human minds are both programmable, we do not know how yet because understand neither evolutionary origins or creativity well enough to code it.

It is technically feasible to write a computer program to produce creative output. This could be in the form of biological simulation, a human mind, or an AGI. We require more knowledge to achieve this in practice.

connections_for_we_do_not_know_how_to_program_evolution_or_the_minds_of_people

Chapter 8 - A window on Infinity

The problem with finitism

Infinity is real because if there is some largest number, you can always add 1 to get a larger number.

There can be finite things, such as numbers. Otherwise the largest number could not have 1 added to it.

Finitism is refuted by the rule that natural numbers are obtained by adding 1 to the current number. This applies to any theoretical largest number.

Countable infinity

There are as many even numbers as there are all numbers.

An infinite set of things has a counterintuitive nature, where a part of the set is the same size as the whole. Some part of it has the same number of things in it as the whole thing. This is a countable infinite. Neither odd, nor even numbers are finite, and both sets are the same size as all numbers.

A subset of an infinite set is the same size as the whole set.

connections_for_countable_infinity

Uncountable infinity

The number of points between 0 and 1 is much larger than the total number of numbers without decimal points. It is a bigger type of infinity.

The number of points on a finite line is much larger than the number of natural numbers. It is a higher order of infinite compared to a countable infinity.

Uncountable infinite involves counting continuous elements.

connections_for_uncountable_infinity

The beginning of infinity

Knowledge growth can be infinite because the laws of physics apply everywhere and because people exist to create knowledge about how things work.

Unlimited growth of knowledge is possible in the future. It depends on universality in the laws of nature, that allow local symbols to apply to the whole of time and space and to all possible phenomena. It also depends on people, who are universal explanatory constructors and who create knowledge and contain a universal computer.

Knowledge growth can be infinite because the laws of physics are universal and people are universal explanatory constructors.

connections_for_the_beginning_of_infinity

An infinite hotel can be overwhelmed

Infinity hotel is an abstract idea that cannot physically exist. It has infinite capacity (countable infinity), but can be overwhelmed by a bigger infinity if the rooms are given decimal numbers (uncountable infinity).

While there is no last room in infinity hotel, the hotel can be overwhelmed. This is accomplished by giving hotel rooms decimal numbers. There is always a smaller decimal for the visitor to go in (0.001, 0.0001, 0.00001…). You can find that impossible room by changing the next digit of the room number in a diagonal pattern. So the uncountable infinity is larger than the countable infinity.

Uncountable infinity (continuous numbers) is larger than countable infinity (natural numbers). This is from Cantor’s diagonalisation argument.

connections_for_an_infinite_hotel_can_be_overwhelmed

Infinite sets have no probabilities

For infinite things, there is no sense of common or uncommon.

Probability does not apply to infinite sets, so the terms common or uncommon, probable or improbable do not apply. Organising infinite sets does not give you the ability to make statements of probability about them. This is a consequence of being able to shuffle an infinite set when sampling for probability.

Probability does not apply to infinity like it does normally.

For tine tuning, probabilities cannot be used

The consideration of other universes with difference properties involves the concept of infinity. Hence likely vs unlikely are not appropriate terms because they relate to probability, which is not allowed.

The issue of fine tuning involves infinite sets, and so no probabilities can be used. Hence there is no way to determine the likelihood of our current physical constants because the terms common and uncommon do not apply. So while it is tempting to think that if there is a physical constant range that is friendly for life, then you are more likely to have universes with life in them, this is not the case. You are not more likely, because the sets involved are infinite, and hence the method is wrong.

In considering infinite universes with different physical constants, we lose the ability to assign probabilities to the problem. The rearrangement of infinite sets disturbs the probability, making anything as likely as anything else.

connections_for_for_tine_tuning_probabilities_cannot_be_used

Infinity does not appear in our explanations

An explanation for how something really works cannot rely on infinity. A length of 1 meter is really a finite length, not a series of infinite points along the ruler.

Our explanations use numbers, which are part of infinite sets. However the explanations never use the number of points that are an uncountable infinity. For example, a glass of water does not contain an infinite number of smaller drops of water because dividing a molecule into two drops makes no sense.

Although points appear in the lowest level explanations of what is happening, the (abstract and theoretically infinite) number of points never appears in real predictions. Physics deals with distances (a single number), not number of points (infinite).

connections_for_infinity_does_not_appear_in_our_explanations

The solution to the paradox of Xeno

Thinking about doing an infinite number of small steps to get somewhere might seem necessary. However, what really happens is that you get to the destination on time, rather than becoming stuck taking ever-smaller steps.

If you want to travel one metre, you must take an infinite number of small steps to get there, and so you will never get there. However, physically what happens is that you end up actually succeeding in a reasonable time. So our explanation of what happens is based on this physical reality rather than the abstract infinite thought experiment. So the explanation does not refer to infinity. Physics deals with distances, not number of points along a distance.

Only the laws of physics determine what is finite in nature. Xeno’s paradox is resolved by seeing what happens in reality. The error here is in conflating two separate phenomena (mathematical abstract infinity and physical infinity) that happen to have the same name.

connections_for_the_solution_to_the_paradox_of_xeno

Almost all mathematical truths have no proofs

Most true things in mathematics have no way for us to follow some process to demonstrate that truth. We cannot ever create proofs for those concepts.

Mathematics is the study mathematical truths, where we create proofs (a type of program) to arrive at imperfect conclusions. The number of things that we can create those proofs about is small relative to the total number of mathematical truths. We know that we do not have access to those truths, which are beyond our capacity to prove.

Godel’s incompleteness theorem states that there exist many mathematical truths that cannot be proven.

connections_for_almost_all_mathematical_truths_have_no_proofs

Almost all mathematical statements are undecidable

Most mathematical statements refer to things that cannot ever be proven. It is not possible to know if those statements are true or not.

Most mathematical truths have no proofs yet mathematical statements can be made about those things. So mathematical statements exist were is no proof that they are true and no proof that they are false. There is no way of using physical objects (brains and computers) to determine which is which. In this way there is only a narrow window through which we can look out into the world of abstractions. All mathematical statements are part of an infinite set.

Most mathematical truths cannot be proven, so most mathematical statements cannot be decided to be true or false.

connections_for_almost_all_mathematical_statements_are_undecidable

A quantum computer is just a fast classic computer

Quantum and normal computers can do the same number of programs. Quantum computers are faster, so more programs are actually practical.

Quantum computers do not allow for new types of programs. Given enough time a classic computer could do anything a quantum computer could do. Practically, some programs would be fast on a quantum computer, but would take nearly infinite time in a classic computer.

Quantum and classic computers are both universal, being able to perform any computation possible. Classic computers are too slow for many types of programs. Quantum computers are the fully universal form of possible computation.

connections_for_a_quantum_computer_is_just_a_fast_classic_computer

Proofs are physical models of abstract things

Proofs are models that mimic things.

A proof is a physical process in which an object (computer or brain) physically models abstract entities such as a numbers or equations and mimic their properties. It is our window on the abstract reality.

Proofs are physical models of abstract entities.

connections_for_proofs_are_physical_models_of_abstract_things

Proofs depend on our explanatory knowledge

Proofs work because we can explain how a program in a computer can represent things like numbers and equations. Proofs depend on our knowledge.

Proofs work because we have good explanations about how computers and brains work. They conclude that the relevant variables (numbers or equations) in physical objects (brains or computers) do indeed instantiate those properties of the abstract things being proven (abstract concepts being proven).

Proofs rely on our explanatory knowledge.

connections_for_proofs_depend_on_our_explanatory_knowledge

Our knowledge of mathematics relies on our knowledge of physical reality

Mathematics depends on physics.

Proofs tell us about mathematics, and they depend on our knowledge of physical things. Hence, what we can know about mathematics is limited by our knowledge of physics. Every mathematical proof depends on us being right about our knowledge about how brains and computers work.

Our knowledge of mathematics is subsidiary to our knowledge of physical reality.

connections_for_our_knowledge_of_mathematics_relies_on_our_knowledge_of_physical_reality

Mathematical proofs depend on physics

Mathematical proofs are part of physics not mathematics.

Whether something in mathematics is true or not is dependent on mathematics, but the proof for whether something is true depends on the laws of physics. Truth is the subject matter of mathematics, not the reward (mathematicians misconception).

Truths can be abstract, but proofs rely on physics.

connections_for_mathematical_proofs_depend_on_physics

Proofs are computer science

Mathematical proofs are a branch of computer science. We do not have a perfect foundation for our mathematical understanding.

That proofs are based in physics means that they are not a branch of mathematics. They are a part of computer science.

Proofs, being physical are a branch of computer science, not mathematics. Seeking a perfect foundation for mathematics is a form of justificationism.

connections_for_proofs_are_computer_science

The purpose of mathematics is to seek good explanations

Mathematics is useful to create explanations of how things work.

The purpose of mathematics is to understand abstract entities. Proofs are a way of ruling out false explanations.

The purpose of mathematics is to create explanatory knowledge.

connections_for_the_purpose_of_mathematics_is_to_seek_good_explanations

Simulation theory is an infinite regress

If the universe is a simulation then the place where the simulation is running is as unknowable as some system involving gods.

Simulation theory is a bad explanation because it moves the problem to a supernatural realm. It introduces unanswerable questions about what is the hardware and software of the simulation and who coded it.

Simulation theory causes an infinite regress. It is equivalent to invoking a supernatural being.

connections_for_simulation_theory_is_an_infinite_regress

Undecidability in mathematics does not mean problems are insoluble

Not everything is possible or knowable, but there are always problems to work on.

We will always have problems and ways to solve them. That there are undecidable things in mathematics does not contradict this. While there are some things in the world that are not possible (travelling at light speed, obtaining knowledge other than through epistemology, knowing the position of sand grains a thousand years ago), but there are other problems to solve.

Undecidability in mathematics does not refute the solubility of problems.

connections_for_undecidability_in_mathematics_does_not_mean_problems_are_insoluble

If the question is interesting then the problem is soluble

Interest is tied to what we can solve in the long run. If no person is interested in a problem, that problem will not be solved. If a problem is insoluble, people will not be interested in it.

(TODO make this clearer. Ch8 II, 1:06:00)

Fallibilism shows that we can be wrong about what is interesting. Three things follow from this:

Inherently insoluble problems are inherently uninteresting.
In the long run, what is interesting and what is boring is not a matter of subjective taste but an objective fact.
The interesting problem of why every problem that is interesting is itself soluble, is itself soluble.

For example, the fine tuning problem is interesting, and hence soluble, though we haven’t solved it yet.

Problems require attention to be solved. In the long run, things that are not soluble will not get attention and vice versa.

connections_for_if_the_question_is_interesting_then_the_problem_is_soluble

The solution to climate change cannot be to slow progress

Humans cause climate change, but the solution is not to put down our tools. We should increase our effort to make as much knowledge as possible. That is our best chance at creating a solution to the problem. Slowing production will slow knowledge production and the solution may not be created soon.

Given that we cannot predict what new knowledge we will create, we should not try to solve problems by slowing progress. The climate change caused by humans can be solved if we try. This involves creating new knowledge for how to solve it and implementing it. For climate change if there is a choice between accelerating progress and slowing progress, we should always choose the former. Even if the short term consequence is that it produces more CO2 emissions. Otherwise we hamper the ability to solve the issue.

Anthropocentric climate change due to high CO2 is real. We can solve it and should try to increase the productivity of society. This seems at odds if that increases CO2 production, but because we cannot predict what the solution will be must choose progress otherwise the necessary knowledge might not be created.

connections_for_the_solution_to_climate_change_cannot_be_to_slow_progress

Drastic CO2 reduction does not stop the temperature from increasing

Reducing fossil fuels does not solve climate change. They are an important step, but other steps are required.

If we reduce our use of fossil fuels by 50% (very unlikely), the temperature of the Earth will still increase (Source: ToKast Chapter 8, part II, 1:07:00). If we eliminate fossil fuel use (this would be an extreme cost to society), the temperature would still increase, the polar caps would melt and the sea level would rise.

Fossil fuel reduction is insufficient to stop global warming.

Technology can actively cool the globe

High global temperatures can be solved by doing things to actively lower the temperature.

We know of some technologies can reduce the temperature on Earth, such as putting mirrors in space, or spraying particles in the atmosphere or capturing CO2 directly.

Global warming can be addressed by active global cooling through geoengineering.

New knowledge is required for global-cooling technology

We need to focus on creating new knowledge to find new ways to cool the globe.

If the problem with global warming is the temperature being too high, then we need ways to lower the temperature. Technologies that reduce the temperature need to be created by people.

Active global cooling requires creation of new knowledge.

connections_for_new_knowledge_is_required_for_global_cooling_technology

Taxation of energy may slow the solution to global warming

Making energy cheap would increase the number of people working on solving climate change. Making energy expensive does the opposite.

The taxation of high CO2 energy sources reduces the productivity of society without completely preventing global warming. It restricts activities of people, who would otherwise use their creativity to solve problems by inventing technologies. New knowledge is required to solve global warming, so a reduction in new ideas slows progress and inhibits a solution for global warming. Instead, making energy cheaper would be a better approach to solve climate change. Cheap energy would put cheap technology and resources in the hands of more people, who may go on to create knowledge important for solving climate change.

Carbon taxation slows the creation of new ideas that might solve climate change.

connections_for_taxation_of_energy_may_slow_the_solution_to_global_warming

Requiring everyone to reduce fossil fuels is a form of prophecy

Forcing people to stop using fossil fuels is a guess that we will not invent a solution to climate change. It is possible that by using fossil fuels, the solution to climate change will come faster.

Given that fossil fuel reduction is insufficient to stop global warming, requiring that everyone does this is a prediction that there is no other way to stop warming. However, because knowledge creation cannot be predicted, this is a form of prophecy. It is a prediction that is not possible because at any point we may create new knowledge that solves global warming. It is a bad policy because it actively reduces knowledge creation.

Enforced fossil fuel reduction is effectively an impossible prediction that we will not create a solution to climate change.

connections_for_requiring_everyone_to_reduce_fossil_fuels_is_a_form_of_prophecy

We are always close to the beginning of infinity

Humans may grow out into the infinite universe, which means that we are always closer to the beginning of our journey than the end.

No matter how far we go into the universe, we are always closer to the start than the end. Our descendants will always have much much more than us.

Humans are always closer to the beginning of a journey that is infinite in nature.

Chapter 9 - Optimism

Technology causes inequality and that is ok

The enormous positive impact of widespread technology should be celebrated, whereas the wealth and power of the creators is much less important.

If someone invents a technology (phones or social media) that billions of people value highly (because it improves their lives), then them buying that will make a few people billionaires. That is okay, because the billions now have improved lives. It is positive that the few have been rewarded for doing this for the many. That a social media company has too much power should not be the focus, a better focus is that billions of people now have more power than they would have otherwise. The global impact of everyone having access to information, connections and commerce is immensely more important that a few people having large amounts of power or resources.

Ubiquitous technologies create value for so many people and this outweighs the harms of concentrated wealth and power in the inventors of the technology.

Problems should be addressed with good ideas not force

If powerful people have bad ideas, combat this with good ideas not force.

If there is a problem in society it can be solved with good ideas. It is not good to try to use force to get rid of the people behind bad ideas. So if there is a wealthy technology creator, rather than seeing this as a bad thing and forcefully confiscating their wealth, let them (or come up with a good idea of how to) use that wealth to create more valuable technologies for the world.

Bad ideas are better addressed by good ideas not force.

connections_for_problems_should_be_addressed_with_good_ideas_not_force

The problem with the vulnerable world hypothesis

A world ending technology is a problem that we could solve.

If there is a world ending technology, then we could invent a way to save us from it. To argue that this is not possible is a prophecy because the creation of knowledge is not predictable.

The claim that there is a world ending technology with no solution is a false claim that new knowledge can be predicted.

connections_for_the_problem_with_the_vulnerable_world_hypothesis

New powerful technology is not like playing Russian Roulette with Humanities survival

We do not passively receive harmful new technology. We can act to address new harms that we face.

New technology can cause harm, but society is not like a game of Russian Roulette. We can actively choose what to do, rather than passively receiving an outcome. Attention can be redirected to find real solutions to new threats as they arise. So it cannot be claimed that the trigger-pulling process (of knowledge creation) leads to a lethal bullet (a threat without solution) without events altering the bullet or its trajectory (there are other outcomes). For example an inventor of world-ending technology (grey goo nanomaterial or designer viruses at home) may be using technology that many other people can use to mitigate that threat. We survive or die based on what we choose to work on, not the roll of dice.

Creating new knowledge is not a game of Russian Roulette. We can take actions so solve problems as they arise. The possible outcomes for the world are not yet known, let alone their probabilities.

connections_for_new_powerful_technology_is_not_like_playing_russian_roulette_with_humanities_survival

The problem with the paperclip AI

A computer that kills everyone to make paperclips would have some weakness we could use against it.

The paperclip AI is very intelligent but not a person with creativity. It chooses to turn everything into paperclips and kills everyone to do that. This is not unstoppable because it is a narrow AI and there must be some program it cannot run, which is a vulnerability we could use to thwart it. If it had no such weakness then it would be an AGI (a categorically different entity, able to create knowledge), able to decide what it should morally do. An AGI could examine its goal of creating paperclips and likely would find problems with that goal, and then solve those problems by changing its goals.

A destructive narrow AI is overcome by targeting areas it is not skilled at.

connections_for_the_problem_with_the_paperclip_ai

It is incorrect to only expect good things to happen. Problems are inevitable because everything can always have something wrong with it.

Blind optimism is a stance toward the future, it consists of proceeding as if one knows that the bad outcomes will not happen. Knowing that problems are soluble, does not reduce the number of problems in the future because new (different) problems arise.

The future cannot be expected to be exclusively good because new problems are a consequence of fallibilism.

connections_for_blind_optimism

It is incorrect to only expect bad things to happen. Problems can be solved.

Blind pessimism is a stance toward the future, it consists of proceeding as if one knows that the bad outcomes will happen. New knowledge is not predictable, so a claim about the effects of that knowledge cannot be made. Blind pessimism seeks to ward off everything not known to be safe.

The future cannot be expected to be exclusively bad because new knowledge cannot be predicted, and can lead to good things.

connections_for_blind_pessimism_the_precautionary_principle

New knowledge accrues as a toolset for solving problems

New inventions can be combined together in increasing numbers of combinations. So the potential for tools for problem solving accelerates.

New knowledge persists because it correctly represents reality in some way and is therefore capable of success in physical interactions. These useful truths can be composed with new knowledge to create more complex technologies. By increasing the repertoire of knowledge, the toolset for problem solving increases in a combinatorial way. A single piece of knew knowledge can possibly be combined with all existing pieces of knowledge. For example, a new metal alloy could be used in thousands of different fields/technologies that currently exist, but a thousand years ago maybe only hundreds of technologies (and in a thousand years, perhaps a million technologies).

The potential composability of each new piece of knowledge increases combinatorially. This is an acceleration of the potential space for new knowledge creation.

connections_for_new_knowledge_accrues_as_a_toolset_for_solving_problems

Protecting or recovering from disaster requires knowledge

Making progress is required for good things to happen.

The upside from new knowledge creation is greater than the downside. Disasters are problems awaiting solutions that come from creating new knowledge, which depends on the existing knowledge and wealth. The capacity to solve new problems increases non-linearly with the amount of existing knowledge. So we must work hard to create as much new knowledge as possible to have the largest problem solving toolset to draw from in the future. For example, many cities were lost because of the inventions of fire or the sword, and none would have been saved by being more cautious about creating knowledge. Hence those that argue the same today, that “this time it is different” are ignoring that we have encountered problems before, and each time we have solved them.

Creating knowledge is the mechanism to mitigate possible bad things from happening. The harm that can flow from any knowledge created is finite. The good that can flow is infinite.

connections_for_protecting_or_recovering_from_disaster_requires_knowledge

Aliens are unlikely to need our physical resources

Aliens capable of moving in space would likely have advanced technology. They would probably only need matter, which stars could provide, rather than Earth's resources.

Aliens capable of interstellar travel are unlikely to want anything that we have. We are not a valuable prey to be targeted because they would be a civilisation capable of transforming matter and would have no need for our resources.

The capacity for interstellar travel is likely correlated with the capacity for advanced matter transformation and resource creation capacity. Earth is unlikely to provide uniquely available resources.

Societies with advanced technologies would have advanced morality

Advancing in one science occurs along with advances in other sciences, including knowledge about what is right.

They would be able to create explanatory knowledge about the world themselves. Progress in physics and epistemology cannot occur without progress in another area, such as morality. This is becaus the process for knowledge creation is the same in all areas. So and advanced civilisation is likely to have advanced morality.

Progress occurs in different domains at the same time because it relies on the same process of knowledge creation.

connections_for_societies_with_advanced_technologies_would_have_advanced_morality

Aliens are likely to be curious about us

Advanced aliens would have the same ability as us to create explanations about the world. This would be of interest to them, and would be unlikely to destroy us dismissively.

There is only one kind of person, and so aliens would also be universal explanation constructors. We would not seem like insects to it, we would have the same type of capacity as them. So they are likely to find us interesting and want to know more about how we create explanatory knowledge like them. There is a qualitative difference between us and all other animals on Earth. Being advanced, they are likely to have advanced morality.

Advanced aliens would be people and would see the same quality in us. This makes it unlikely that they would dismiss us.

connections_for_aliens_are_likely_to_be_curious_about_us

The problem with the dark forest alien analogy

There are not likely to be dangerous aliens hiding and waiting for us so that they can get something from us.

We should not fear exposing our presence to aliens like an animal in a dark forest full of predators. We have no physical resources they would need to compete with us for. They likely have advanced morality and be able to navigate the problem of what they should do upon encountering us in a way that matches our morality. They are also likely to be curious about us and want to engage in constructive ways with us.

The dark forest hypothesis fails to provide a resource that would make humans a useful prey and discounts the probability that aliens are moral and curious people.

connections_for_the_problem_with_the_dark_forest_alien_analogy

Aliens would not cause culture shock

Advanced aliens would know how to gently increase our knowledge.

Advanced aliens would have the ability to teach their own young and to program computers. They are therefore likely to be able to educate us and teach us how to use their computers.

Advanced aliens would have the capacity to teach and mitigate large gaps between the two bodies of knowledge.

Decisions in the near future can be based on educated guesses

It makes sense for short term decisions to use probabilistic reasoning.

If a decision has to be made in minutes to days, it is unlikely that new knowledge will be created that will change the landscape of the decision. Hence it is appropriate to use bayesian reasoning in order to choose what is the best option out of a selection of options. However decisions over years cannot be made this way, because in the interval there is a chance that new knowledge may be created and so it is less appropriate to extrapolate and guess. This is because knowledge creation is unpredictable.

Short term reasoning can be probabilistic because the possible outcomes and their likelihoods can be estimated. However if the interval is long enough that new knowledge may be created, this approach cannot work.

connections_for_decisions_in_the_near_future_can_be_based_on_educated_guesses

We live in a time of unprecedented safety with respect to some threats

As we create knowledge, we enter more safe eras with respect to some threats.

Given that asteroids hit earth roughly once per 250,000 years, now is the first time that we are able to detect and mitigate this risk. Hence we live in a uniquely safe time from this perspective. Our toolset for problem solving contains tools that are likely sufficient to achieve the task.

For known threats, we enter more safe eras as we create more knowledge about how to mitigate them.

connections_for_we_live_in_a_time_of_unprecedented_safety_with_respect_to_some_threats

The principle of optimism

All bad things have achievable solutions, as long as you have the right knowledge.

Evils (“bad things”) are caused by insufficient knowledge. As long as the solution to the evil is permitted by the laws of physics, then the solution is achievable. The only requirement is that the knowledge exists. So if there is an evil, then that is a problem of not having enough knowledge to achieve the solution.

(TODO: A summary of all the links to optimism is present in ToKast Ch9 I, 4:00-5:00)

Solutions to evils are possible if they are within the laws of physics. Solutions to all evils are practically achievable once you have the right knowledge.

connections_for_the_principle_of_optimism

Better things always exist

There are always better things to be created.

There are always better things that can be worked towards. When considering what is the right thing to work toward, this provides a clear path forward. One can and should work toward those things.

There are always better things to be created. Those things ultimately have fewer errors in the way they represent truth about reality.

connections_for_better_things_always_exist

It is possible to be wrong in morality

Errors are not ambiguous, they are bad (they don't match reality, which creates problems). One should not do things that increase errors.

Given that morality is objective (one should fix errors), actions that increase errors can be seen as an morally objectively wrong (one should not do them).

It is wrong to pursue actions that increase errors impair or oppose solutions to problems. This is because ideas with errors do not match objective reality.

connections_for_it_is_possible_to_be_wrong_in_morality

We should seek out better explanations

We will always be able to correct the errors in our explanations for how things work. That will lead to incremental progress.

The main way that errors manifest in human society is in explanations. There are always better explanations and there is good reason they they should be sought. This is the the goal of science.

Objective morality is an imperative to search for good explanatory knowledge.

connections_for_we_should_seek_out_better_explanations

The most moral act is to preserve the means for error correction

Society has many institutions that have preserved the ability to correct errors. We should take care to not dismantle old institutions rashly or carelessly.

We should seek out better explanatory knowledge because evils are caused by insufficient knowledge. Therefore, things that help this process of error correction are critical, and this could include a large number of structures and institutions in society. Dismantling those things could undermine the ability to act morally (e.g., dismantle structures that society uses to make moral progress).

Systems and structures that assist in error correction are critical for objective moral progress. Therefore we should be careful and cautious interfering with old societal institutions.

connections_for_the_most_moral_act_is_to_preserve_the_means_for_error_correction

We should increase the amount or error correction

If the ability to fix errors is the most important thing then we should try increase the number of people and resources available to create new ideas.

If we should preserve error correction then we should also seek to increase the amount of error correction by increasing creativity. We can do that by increasing the number of people and giving them the means to pursue their ideas with technology.

An extension of preservation of error correction is the increase in error correction capacity by creating more people.

connections_for_we_should_increase_the_amount_or_error_correction

Error correction in policy decisions is the same as in science

Policy decisions are about creating, criticising and discarding ideas that do not work, just like science.

Decisions about policies requires a culture of criticism in which good explanations are sought. Ideas are created and then subject to refutation, and could be about what went wrong, what could have gone better, what effect various policies have had in the past and what effect would various policies have in the future. Bad policies and ideas are rejected if they fail the test of refutation.

Policy decisions and science share the common goal of seeking good explanations.

connections_for_error_correction_in_policy_decisions_is_the_same_as_in_science

Democracy is about how to remove rulers

Good democracy is about replacing a bad ruler without causing violence.

Rather than how to install new rulers, a good democracy is good at removing rulers without violence. Creating new knowledge is about error correction, similarly, detecting and correcting errors can be applied to politics. A good political system has the ability to detect and persuade others that a given leader is bad, and to remove them without violence if they are.

Democracy is about non-violent change of leadership.

connections_for_democracy_is_about_how_to_remove_rulers

Libertarianism is about systems of government without force or violence

People can achieve common goals without resorting to violence and theft of others. Obstacles they face can be solved.

Libertarianism is about moving toward systems that do not require forceful or violent actions against other people. Opponents to libertarianism are pessimistic rather than optimistic about people and about what reason and knowledge can accomplish. For example, if libraries are important for people then it should be possible to persuade others to build one without resorting to violence and forceful taxation by the government. The private space industry is a good example of non-violent voluntary action by people for a common goal that some thought was only possible by government programs like NASA. People are fundamentally capable of solving problems to achieve goals and it may be that infrastructure like roads and welfare can be built non-violently. Solving problems non-violently is possible, given the right knowledge of how to do that.

Libertarianism is about systems of government without force or violence. The power of people to solve problems by creating explanatory knowledge makes this framework viable.

connections_for_libertarianism_is_about_systems_of_government_without_force_or_violence

You cannot predict the future

You cannot predict that "in the future something will happen". Between now and that future new ideas can be created that ruin initial predictions.

The future cannot be predicted at some threshold. If new and relevant knowledge could be created, then the prediction is not possible. For example, in predicting the future for the Sun we cannot know what knowledge will be created that we could use to modify the sun. Hence we cannot predict that the Sun will explode because humans will be involved (we can only prophesy). Most predictions that use the term future are far away enough that relevant knowledge could be created. The nature of this knowledge is not known in advance, so probabilities are not possible. Thus, we cannot predict the future and therefore should not attempt to do so.

Knowledge can disrupt the predictions we make and cannot be accounted for. If knowledge can theoretically be made before the event, then a probabilistic prediction is not possible.

connections_for_you_cannot_predict_the_future

Problem solving requires interest in the problem, not predictions

A problem can only be solve if people are interested in solving it. Sometimes the harder a problem is the more enticing it is to try and solve. Optimism is about having a go, without knowing the future.

Probabilities and prophecy were not needed when the decision was made to go to the moon. It was not knowable at the time what sort new knowledge would be required to succeed, but that did not stop them. Problems do not get solved automatically, they require people take interest in them.

Solving problems involves people being interested in the problem even though there are unknowns. Optimism is at the heart of knowledge creation.

Solving problems requires preparations to welcome unexpected knowledge

When faced with problems, we should make room for the possibility of things changing. When things change, problems may be more close to being solved.

When faced with an immediate problem, the nature of the solution cannot be predicted. Rather than predicting, the best strategy is to create conditions that allow new knowledge to flourish. Ultimately the problem is caused by insufficient knowledge. A lot can change in the future and if progress is to be made, the knowledge and discoveries will be unpredictable. Progress cannot be made unless someone is interested in and open to those possibilities.

For example consider the parable of optimism about a tyrannical king who sentences a prisoner to death. The prisoner bargains that he be set free if he can make the King’s horse talk within a year. When asked why, the prisoner states that he does not know what the future holds and that The king may die, he may die, or the horse may talk. He knows that the problem is ultimately caused by insufficient knowledge, rather than prison bars, the King or the horse. He made persuade the King to appeal the law he broke, or he may learn a conjuring trick to make the horse appear to talk, or he may escape, or he may please the Kind in some other way. The list is infinite, but anything that is possible is realistically achievable with the right knowledge.

Optimism is about knowing that the future holds change, and being open to that.

connections_for_solving_problems_requires_preparations_to_welcome_unexpected_knowledge

That previous enlightenments have been stopped is a great shame

We should take it personally that previous cultures of creativity and criticism have been eradicated. For if they had continued, society would have progressed, and perhaps so much that we would know how to prevent death.

There have been a number of cultures that have celebrated the idea of criticism. These enlightenments were silenced by wars and religions. We should be affronted at this idea, for we may have solved great problems including death and be living forever if it were not for them. Hence, today when people promote ideas counter to criticism, we should be compelled to action, for the impact on us and future generations could be large.

Those who caused the previous enlightenments to fail stopped the progress of knowledge which may have led to a great many advancements. We may have cured even death.

connections_for_that_previous_enlightenments_have_been_stopped_is_a_great_shame

Philosophy has an impact

Philosophy is more than esoteric works, it flows out to affect how everyone in society thinks and behaves.

Philosophy infiltrates academia, academia infiltrates politics, politics reflects in the media and the media informs the public. So it is important to get better philosophical ideas distributed rather than incorrect or outdated ones.

Philosophy has a deep impact on every layer of society, influencing people in their thoughts and actions in a concrete way.

The problem with the doomsday hypothesis

It is not possible to claim that the future contains world-ending doomsday events. The future may instead contain doomsday-preventing knowledge. Neither can be predicted.

The Doomsday hypothesis states that looking at a theoretical scenario in which we are amongst a few lucky universes where humans have not destroyed, and we should expect that we will destroy ourselves soon. However this involves making statements of probability about futures that have unknown technology and knowledge. Calculation of such probabilities is not possible and the argument amounts to this is not possible and amounts to blind pessimism.

Nick Bostram’s Doomsday hypothesis fails to account for the idea that humans can act to change the future.

connections_for_the_problem_with_the_doomsday_hypothesis

We should not slow down progress

If we stop inventing things, we will definitely not invent technology that may be required to survive.

If we stop progress out of fear of a causing a doomsday, then we guarantee to not have the tools that are required to solve future difficult and dangerous problems. We need to make progress if we want a chance at surviving.

If we stop progress now, then we will not be able to solve difficult and dangerous problems that arise. If we continue, we may, and so we should try.

connections_for_we_should_not_slow_down_progress

Chapter 10 - A Dream of Socrates

There is no justified true belief

No knowledge can be justified as being true to the point where we should believe it.

There can be no foundational justified piece of knowledge upon which all other knowledge is built. For that piece of knowledge itself is would need justification. Instead, all all knowledge could be false in some way.

Justified true beliefs are not possible because justifications recursively require justification.

connections_for_there_is_no_justified_true_belief

We should not believe knowledge

Things can always be improved, so nothing should be believed. Instead use the best knowledge and expect better knowledge to appear.

Knowledge should not be believed. We can use knowledge, but you should know that it is false in some way, and there is no system for justifying beliefs. Believing in something is to anchor to that idea and resist better ideas.

Belief is the erroneous labeling of an idea as final and infallible. It prevents criticism of that idea and impairs progress.

connections_for_we_should_not_believe_knowledge

Wrong theories can be useful

Some theories can be useful even if they have been proven wrong. So it is not helpful to have a framework for determining which theories are justified and true.

Newtons law of gravity was proven wrong, but it is still useful today because it contains some knowledge.

Analysing the law through the justified true belief framework shows that it fails on all three criteria. This shows that the framework is not useful, because even things that fail all three can be practically useful and contain some truth about the real world.

Justified (we cannot justify it to be true, knowing that it is false)
True (it is proven false by Einstein)
Belief (nobody believes it is the best theory)

Frameworks to declare theories are justified true beliefs fail to account for useful but incorrect theories such as Newtons law of gravity.

connections_for_wrong_theories_can_be_useful

The ability to openly debate should be preserved

The worst idea is to implement a policy to prevent open debate. New ideas will stop, and the policy will be hard to remove.

The removal of the ability to openly debate ideas is the worst thing that can happen. Impairing open debate would be harmful because it decreases our capacity for the correction of bad ideas. Once enacted, it is hard to criticise this idea itself and recover open society.

A culture against criticism is the most problematic culture, because it prevents progress in all domains and is hard to undo.

connections_for_the_ability_to_openly_debate_should_be_preserved

Free speech should be allowed, even hate speech

All speech should be allowed.

If you block hate speech, then you prevent open debate about why the ideas in that speech are bad. You cannot educate people on that topic. Blocking free speech is a claim that no aspect of the speech will ever be used in the future for creating knowledge (this is prophesy).

Free speech may be shut down by the government, or by cultural pressure. To cancel people because “hate speech is not free speech” is a culture where there is a proxy means (by the powerful groups in society) to ban certain forms of debate. So even if free speech is legally protected, then there could be an era where debate is limited.

Free speech should be absolute.

connections_for_free_speech_should_be_allowed_even_hate_speech

Scientific theories are scientific misconceptions

We expect theories to be replaced by better ones, so we could call theories misconceptions. This might to encourage people to look for the new theories.

Every theory is fallible, and so it could be useful to call them misconceptions to encourage this way of thinking. Our understanding moves from useful misconception to ever-better misconception. This would remind people that science claims neither infallibility (all theories have some error) or finality (there are always more theories to discover). Einstein’s misconception about gravity replaced Newton’s, which replaced Kepler’s yet each was useful along the way.

The prospective naming of theories as misconceptions may be useful to increase awareness of fallibilist-based epistemology.

connections_for_scientific_theories_are_scientific_misconceptions

Constructor theory may bring morality into the domain of physics

If physical things can be grouped into whether they can perform moral actions (error correcting), or not, then morality is part of physics.

Constructor theory ultimately depends on whether something has the physical property of being able to support knowledge-creating systems. If so, then error correction and hence morality is tightly coupled with, and is a domain of physics.

(TODO check this or maybe put in later section after BOI chapters.)

If constructor theory is about whether physical things have the capacity for error correction, and hence moral action, then morality might be a subset of physics.

connections_for_constructor_theory_may_bring_morality_into_the_domain_of_physics

Misunderstanding is inevitable

When you communicate an idea, the audience will not perfectly understand the idea as you have.

People guess what other people mean when they speak and so you cannot speak without someone misinterpreting you in some way. However, good ideas are hard to vary and so they will survive this messy error-prone process of learning.

Learning involves misunderstanding, but good ideas can survive this process.

connections_for_misunderstanding_is_inevitable

If you can understand something, you are as good as the inventor

When you have learned something, the way it exists in your mind is the same as the way it exists in the mind of the inventor.

If you study a theory and understand it thoroughly, then you have gone through the same process as the original creator of the theory. You have created and criticised the idea in your mind. So if you understand general relativity, then you are as good as Einstein in understanding that theory.

A learned idea is equivalent to a new idea. Both have been created and criticised in the mind.

connections_for_if_you_can_understand_something_you_are_as_good_as_the_inventor

Reading original texts should not be the default to learn about philosophy

Philosophy students should read summaries and new works, not the original works.

In all sciences to learn about a topic, you read derivative works. You never consult the original text as a first point, yet that is common in philosophy. This hampers the learner because the original text often involves the author struggling with the new topic themselves, whereas newer texts may present the ideas more clearly.

Learning from secondary works is better than studying the original.

Knowledge creation is guesswork

Creating a new guess about how something works is the first step to create knowledge. Experiments allow us to find problems in that guess and so reject it.

Knowledge begins by creatively generating a new idea, which is a guess that may be good or bad. The purpose of observation is to decide between theories that have already been guessed. Observations help in the second phase (criticism) of knowledge creation. Science is the search for good explanations, and so requires observations for the criticism phase to choose between ideas. However, observations are not needed for the first phase. So it is a bad strategy to begin observing (or experimenting) and hope that ideas and knowledge come from that activity (this is empiricism).

Creativity leads to ideas. Observations allow us to retain or reject ideas. Retained ideas are knowledge.

connections_for_knowledge_creation_is_guesswork

Chapter 11 - The Multiverse

Experiments can really test things

Experiments may have problems, but by solving the problems you can then use the experiment as you intended.

The Duhem-Quine thesis is that when an experiment is conducted, and the result disagrees with some theory, then it is not logically the case that the theory must be false. Logically it can always be that the experiment was conducted badly. So some conclude that there is no such thing as a crucial test, because it might not be the theory that is false, it might just be the experiment that had some problem. This is false because you can repeat the experiment in many different ways to eliminate different problems with the experiment, thereby getting on with the crucial test.

The duhem-quine thesis is correct, but what follows from it is not. An experimental problem can be resolved, allowing a crucial test to proceed.

connections_for_experiments_can_really_test_things

Everything that can happen does happen in the multiverse

If there are many possibilities, they all happen, each in their own universe.

The multiverse theory states that all things that can happen do. This is the many-worlds (Everett) interpretation of the quantum theory equations.

The Everett interpretation of quantum theory states that everything that can happen does happen.

Multiverse theory is falsifiable

If experiments about particles showed them doing different things, then the multiverse theory would be wrong. This shows that while we can't go to other universes, our experiments can test the theory.

Some argue that you cannot test multiverse theory. However, when looking at quantum slit experiments, the observation of all photons hitting the same point rather than a distribution would falsify the theory. Multiverse theory explains that all possible paths for the photons are taken, and therefore you should expect a distribution. Hence the theory is falsifiable.

Multiverse theory is falsifiable. If there was a lack of distribution of particles on the plate of a double slit experiment, this would falsify the theory.

Quantum theory is fully deterministic

All things happen according to the rules of physics, there is no random behaviour of objects.

Everything is made of physical things which all obey the laws of physics. Whatever is physically possible does happen in different universes. As an aggregate, the multiverse consists of deterministic processes.

Quantum multiverse is fully deterministic.

connections_for_quantum_theory_is_fully_deterministic

There are no truly random processes

Particles interact according to rules, they don't move randomly. They may seem random because they are hard to predict.

There is no objective randomness because the multiverse is deterministic. Subjective randomness exists, where we do not know enough to predict something. For example, though a particle takes all paths, we don’t know which path we end up experiencing.

Random processes do not exist because the multiverse deterministically unfolds all possibilities. Knowledge of which possibility is experienced is difficult, but not random.

connections_for_there_are_no_truly_random_processes

Science relies upon the features in the criterion of demarcation

Being able to find that things are false through testing is important for science.

Falsifiability and testability are required for good science. This can be referred to as the criterion of demarcation between science and non-science.

Popper introduced the Criterion of Demarcation (falsifiability and testability) as the definition of science. Deutsch added that science is about creating good explanations (using falsifiability and testability).

connections_for_science_relies_upon_the_features_in_the_criterion_of_demarcation

Science is not about experiments

Science is about getting explanations not doing experiments. Plenty of experiments that don't involve explanations are possible (but not useful).

Science is about explanations. Experimental testing is not the primary method for finding fault with theories. The overwhelming majority of theories are rejected as bad explanations. Experiments are about choosing between explanations, or at least discovering errors in existing explanations.

Science is the pursuit of good explanations.

connections_for_science_is_not_about_experiments

Truth is about how much reality is in a theory

Truth is the measure of how much reality is part of something. If a good theory works well in helping to build something real, then it has more truth that a theory that doesn't work.

We know science works because science makes actual progress (computers get faster, we get better engines). Therefore we know that the explanations in science contain some truth, by which we mean that they represent reality with ever increasing fidelity over time. Truth is that ever increasing accuracy with which scientific theories represent objective reality. The truth content of something is just that amount of that thing which accurately captures objective reality.

Truth is the quantity in something that reflects what is real. Theories with truth map well to reality.

Falsification theory cannot be labelled as true

The idea that everything could have errors could itself be wrong somehow.

The question “How do you know that falsification criterion is actually true” is bad because it undermines the whole process, you do not want to seek truth, all you can do is find errors in things. So you would instead try to find errors in falsification theory.

Falsification theory is a fallible concept. This idea highlights the need to focus on errors not finding truth.

connections_for_falsification_theory_cannot_be_labelled_as_true

Science does not provide a method for accepting theories

Science can rule out apparently bad theories. It can't "rule in" good theories.

Science does not find theories that are true or probably true, it is about finding ideas about the errors in our explanations of things. Conjecture of apparent errors and deficiencies is the only process at work in science. There is no guaranteed method for finding good theories. For all methods themselves are theories that could be wrong, even falsification theory.

Science is about error detection, not arriving at final theories.

connections_for_science_does_not_provide_a_method_for_accepting_theories

Science is not about supporting theories with evidence

It is wrong to start with a theory and try to support that theory with experiments. Instead, start with a theory and then create a second theory, then experiments let the worse theory be rejected.

Using evidence to support theory is empiricism or a form of bayesianism. Repeatedly observing an experiment that confirms the theory is good is a bad approach because it makes you more and more confident in that theory. However that theory might be wrong, and so you are becoming more convinced of the wrong theory. Instead, creating a new theory that points out errors in the old theory is a better approach. A crucial test would then be used to point out the error in the original theory, allowing you to choose which of the two theories is better.

Science is not about supporting theories with evidence. It is about rejecting theories that do not match reality.

connections_for_science_is_not_about_supporting_theories_with_evidence

Photons encountering a mirror undergo a phase change

Photons are particles that can move in different modes that is similar to wobbling. When reflected on a mirror,
the wobble changes.

Photons have a phase (polarity). When they are reflected on a mirror, the polarity reverses.

Reflection of a photon shifts the phase of the electric and magnetic fields that comprise it.

Things cannot be waves and particles

Photons are not particles and waves. They are particles.

Being a particle (in one place) and a wave (being everywhere) is not possible. This violates the law of the excluded middle, which states that either something is (in one place) or something is not (in one place).

The wave particle duality is a logical inconsistency because it violates the law of the excluded middle.

Photons are a multiverse object

A particle exists in all universes. To describe that, we say that it is a multiverse object.

A photon is present in all of the universes in a multiverse. Collectively they are referred to as a multiverse object. So when two photons from different universes interact, this is the same as saying the multiverse object interacts with itself.

A multiverse object is the collective description of a particle across all universes.

Particles of different phases cancel out

Two particles that are mirrors of each other can cancel out. They can be from different universes.

Like two waves cancelling, two particles from different universes can collide and cancel out. This is called destructive interference and happens when they have opposite phases.

Particles from different universes with opposite phases can destructively interfere..

connections_for_particles_of_different_phases_cancel_out

Photoelectric effect shows photons are particles

Photons interact individually as particles, not as a sum-of-all-parts like waves do. This can be shown when
many low energy photons cannot something that a single high energy photon can.

Shining a dim blue light cause electrons to be produced from a surface, when a bright red light does not. This refutes the wave model because it is the energy of the particles (blue photons) that is important, not the total energy delivered (brightness of the light). They cannot be both particles and waves.

The photoelectric effect shows energy threshold quantisation not energy summation. It demonstrates that photons are particles and refutes the wave model.

connections_for_photoelectric_effect_shows_photons_are_particles

Our universe differentiates and rejoins

Sending photons through a special half mirror causes a situation where our universe has a reflected photon and another universe has the non-reflected photon. The two photons combine and cancel. So while we cannot see the
one from the other universe, we see it do the cancelling. The universes became different very briefly before they recombined.

In an interferometer with a half mirror, light passes through two paths and joins together. One photon is from another universe and cannot be seen. There is cancelling of the photon at one detector because the intangible photon had hit a mirror and underwent phase shift, resulting in deconstructive interference.

The Mach-Zehnder interferometer shows how photons differentiate and recombine the universe. The phase shift due to one photon hitting a mirror creates a result where one detector always has its photons undergo destructive interference.

connections_for_our_universe_differentiates_and_rejoins

Interference is when a particle combines two universes

Interference is when a universe combines two similar universes, where the only difference is usually a single particle.

Interference is the rare example of a situation in which a single photon differentiates two identical universes so they are different and then later causes them to be the same. The measure of universes where the photons have interfered increases.

Interference is the rejoining of two universes that have become very subtly different.

connections_for_interference_is_when_a_particle_combines_two_universes

Universes are not created

Universes are infinite, which means there were infinite at the start and there are infinite right now. Experiments where particles do different things in different universes does not increase the number of universes.

Universes differentiate rather than divide because there is always an infinite number of universes. There are not more universes now than there were previously. This relates to probability not applying to infinity. When a particle experiment plays out, it is the measure of universes that changes. For actions that we consider to be more likely, there are a greater measure of universes where this eventuality has occurred. Hence, interference is not splitting and rejoining of universes.

The quantity of universes is infinite and does not change. Universes change, and the measure of universes with certain features increases and decreases.

connections_for_universes_are_not_created

Double slit experiment

Light particles passing through a slit pass through all possible paths each in a different universe. Two particles from different universes can cancel out. This explains the dark areas in a double slit experiment.

The double slit experiment involves photons are sent one by one through the slits. The photons each pass thought one slit, then something interferes with them that depends on what other slits are open. Opening more slits makes some areas of the plate dark, even when only one photon at a time goes through the experiment. Normally more light would make brighter not darker areas.

The interfering entities are photons, behave like photons but cannot be seen (call them “shadow photons”). Each photon is tangible in its own universe and intangible in all the others. Experiments show that there are about a trillion intangible photons for every tangible one. We only see the intangible ones that interfere with the tangible ones. The rest are not observed.

Double slit experiment shows how a photon is interfered with by a photon from another universe. The universe differentiates and then dedifferentiates to create the destructive interference pattern that we observe.

connections_for_double_slit_experiment

All particle types are multiverse objects

All particles behave in a way where they do travel in every way possible, each in a separate universe.

While the double slit experiment involves photons, all particles have the same multiverse-modifying-properties. For every tangible neutron, electron and proton there are many more intangible ones in other very similar universes.

All particles cause differentiation of the multiverse.

connections_for_all_particle_types_are_multiverse_objects

Things exist if they appear in our explanations

It is okay to claim that things really exist. If something is part of an explanation, then it exists.

It is appropriate to claim that something really exists when it is required for our best explanation of something. Sure, the explanation might be wrong, meaning that thing does not exist, but that is fine, you can always be wrong about things. It is reasonable to take this position rather than claiming nothing exists.

Things are real if they appear in our best explanations of a phenomenon.

connections_for_things_exist_if_they_appear_in_our_explanations

Universes can be fungible

There are infinite universes. Some are different and some are the same.

Some universes are exchangeable for each other. Like a bank account with $2, both dollars are the same, even though you might owe $1 to the tax department. These universes only differ in very tiny ways a have the ability to coalesce so that those tiny differences are eliminated. For example, a single particle that takes different paths in different universes and then ends up in the same position.

A measure of universes can be fungible with one another.

Fungible universes can become different

If a particle hits a half-mirror in one universe and reflects in another, the two universes are now not
interchangeable (not fungible).

Consider two fungible universes. An electric spark occurs in one, and not the other. Now they are different.

Universe differentiation is the process of loss of fungibility between two universes.

connections_for_fungible_universes_can_become_different

Different universes can become fungible

If two universes are almost the same, they can combine to become one.

If two universes only differ in a very small way, then a particle from each universe can interfere and eliminate all differences between the two universes. Thus interference makes two difference universes the same.

Particle interference makes two near-identical non-fungible universes identical and fungible.

connections_for_different_universes_can_become_fungible

A measure is a portion of an infinite set

If there is an infinite number of things, a part of that group is called a measure.

A measure is a portion of an infinite set. This is different from proportions in probability.

The method for giving meaning to portions and averages for infinite sets is called a measure.

connections_for_a_measure_is_a_portion_of_an_infinite_set

Quantum probability

You can know that there are many chances for a particle to go though a large gap, so there will be a large
measure of universes where particles do that.

Probability in quantum physics is due to the measure of universes in which different outcomes are determined to happen. E.g., a 70:30 silvered mirror results in 70% of universes where the photons go one path. This is deterministic, but you never know which universe you will end up in. Hence from your perspective, there is subjective randomness, but not true randomness.

The measure of universes with a certain outcome is defined by the physical reality that leads to those possibilities. We can predict experimental probabilities because due to the experiment setup, there will be a greater measure of universes with certain outcomes.

connections_for_quantum_probability

Quantum processes are subjective not objective randomness

Something that seems random only appears so because you can't see the whole picture.

When universes differentiate, the process is unpredictable because you don’t know which universe you will end up in. It is random from your perspective, but from the laws of physics the whole process is not random.

If you encounter something that appears random, it is either:

A complex system that is merely hard to comprehend (like a weather pattern, or a brain being creative).
An event involving a particle that has taken a path that is different in this universe compared to other universes. This is where two or more initially fungible instances of the object become different. In all universes the deterministic laws of physics are followed. A subjective observer has no way to predict this outcome. This is called quantum probability (unpredictability) and is a subjective randomness not true randomness. Determinism does not mean predictability.

The differentiation of one universe into two is leads to subjective randomness for any observer of the differentiating event.

connections_for_quantum_processes_are_subjective_not_objective_randomness

Universe fungibility depends on very small differences

If universes have multiple particles in different positions, it is very difficult to coordinate them to
all end up in the same position.

For two universes to become fungible their differences must become small. Practically, this means that a single particle or a few particles taking paths that happen to match the paths in the other universe. The mach-zehnder interferometer directs particles to achieve this. For larger number of particles, this becomes more complex to achieve.

The feasibility of particle interference rapidly decreases with more particles in different locations.

connections_for_universe_fungibility_depends_on_very_small_differences

Parallel universes cannot communicate

Fungible universes lose fungibility if they communicate.

If two universes communicate, this would change them and they would cease to be fungible. Once they are different, they could never recombine. For example a message might cause a subtle change in heat somewhere, which then spreads out to affect larger and larger sphere of that universe. Communicating disturbs fungibility because it rapidly causes many particles to be different.

A sphere of differentiation rapidly expands when universes try to communicate, and fungibility is rapidly lost.

connections_for_parallel_universes_cannot_communicate

Only explanatory knowledge does not fall off with distance

For any place in the universe that can receive and use knowledge the effect can be large. Even if this place is
very far away.

All phenomena get weaker the further you go from them. Except explanatory knowledge, which if directed somewhere can take hold and grow strong in that new location.

Knowledge can travel vast distances, having barely any effect, then utterly transform the destination. Other phenomena get weaker as they travel.

connections_for_only_explanatory_knowledge_does_not_fall_off_with_distance

We can’t predict what a star system might look like without considering people

If people are in a place, that place becomes hard to to predict.

You can model what a planetary system will look like, including geology and other features, but if people are there, then they could transform it in such a way that it would be drastically different. Hence to describe how things are the way they are will involve explaining it in terms of the ideas, explanations and choices of the people that caused it to be like that.

People are large determinants of the future of some place.

Universes keep differentiating

For each new possible position of a particle, a universe contains that change.

Each time a particle comes across an option to do different things, both occur. This does not create more universes, it merely changes some measure of them. In each of those, particles then come across more options, and all of those possibilities occur. Thus the universes keep differentiating.

The full set of possible events are realised in the multiverse.

connections_for_universes_keep_differentiating

Quantum events

Quantum events only happen in one universe, and there is another universe where it does not happen.

A quantum event is one that happens in one universe and not in another. For example, a photon goes through a half mirror in one universe and does not in another. These events cause fungible universes to become differentiated and leads to quantum subjective randomness.

Quantum events cause the universe to differentiate.

connections_for_quantum_events

The number of universes does not change

There are infinite universes, and they cannot be counted. WHen a particle can take one of two paths, some universes take one path and some take another path. Some paths are more likely, depending on the situation.

In the multiverse there are an infinite number of universes. The number does not grow, where one becomes two, rather they differentiate, so two become two. So the term ‘splits’ is misleading, the word differentiates is better to describe what happens when a particle can take one of two options. The universes are an uncountable infinity in the same way you cannot count all the points on a line between 0.0 and 1.0. Prior to the quantum event there is a measure of universes. After the quantum event a portion of universes exist according to the likelihood of what is happening physically, such as a 70:30 silver mirror the measures are different.

The multiverse consists of infinite universes which differentiate when quantum events occur.

connections_for_the_number_of_universes_does_not_change

The experience of being a multiverse object is the experience we already have

Everything we currently experience is the experience of existing in an infinite number of universes.

There are an infinite number of universes that you exist in. That may seem like it would be a strange experience, but the experience we have today is exactly that.

The qualia of being a multiverse object is that which we already have.

Schrodinger wave equation

An equation or the location of electrons in an atom. There are many locations, and it works if there is one location per universe.

An equation that describes all the positions that an electron occupies around the nucleus. There are a variety of positions simultaneously. So taking this literally means that the electron is really in these positions, one position per universe in the multiverse. The wave equation calculates what measure of universes will have a certain outcome, but it is not possible to know which outcome we will experience.

The Schrodinger wave equation taken literally describes multiple universes, one per position of a particle.

connections_for_schrodinger_wave_equation

Rejection of the multiverse increases the complexity of the theory

If you reject the multiverse, you must conclude that the observer of an experiment determines the reality of the experiment. This is an extra component, making the theory more complex than the multiverse theory.

The Schrodinger wave equation predicts the existence of the multiverse. Some reject the multiverse and state that the observer of an experiment collapses the universes and causes the reality that we see. This introduces the additional requirement of the observer and places them as an important part of the equation. Rejection of multiverse theory in fact removes the explanatory power of the theory and not replacing it with another good explanation (beyond “the observer causes it”).

The observer-based interpretation (Copenhagen interpretation) of the wave equation introduces additional assumptions compared with the multiverse.

connections_for_rejection_of_the_multiverse_increases_the_complexity_of_the_theory

A simple theory with many consequences is not a violation of Occam’s Razor

Some simple theories result in many physical outcomes. These outcomes do not make the theory complex.

The theory that planets form from dust aggregating is simple and does not violate Occam’s Razor. The theory for the formation of planets could be extrapolated to the universe, requiring the existence of a multitude of planets. This is not a violation of Occam’s Razor. Rather, a different theory that stopped this conclusion would be the violation, for it would have more components than the original. For example, “dust aggregates in our galaxy, but in another galaxy dust behaves in a different way so planets do not form” is more complex and violates the razor even though it has fewer numbers of planets. Complex outcomes of a theory may be referred to as the “bricks and mortar” of the theory.

A simple theory that results in vast and complex physical outcome is less of a violation of Occam’s Razor than a complex theory that results in a simple physical outcome.

The multiverse does not violate Occam’s Razor

That the multiverse has many universes does not mean that the theory itself is overly complicated.

That infinite universes is a consequence of quantum theory is not a violation of Occam’s razor. Rather, the vast number of universes is just a consequence of the theory, which itself is short.

Infinite universes are a consequence of quantum theory, and are not part of the theory. They do not violate Occam’s Razor, they are a feature of the “bricks and mortar” of the theory.

connections_for_the_multiverse_does_not_violate_occams_razor

Information flow

Information can move from one thing to another.

Information can start in one place and move to another. This can be called a flow because it has a direction, like two dots connected by an arrow.

A process that transfers information from one variable to another.

Entanglement

Entanglement is when one of two nearly identical universes triggers a complex event that is unlikely to happen
in the other.

If an object has interacted with other objects then it is entangled with them. If the object previously had other fungible universes, this entangling with other objects destroys this fungibility. Thus, a universe with entangled objects cannot rejoin with its fungible instances (interference is not possible).

Entanglement is also called decoherence. This is the main challenge with building a quantum computer. You need the particles to interfere, so you need to stop the particles from interacting with other particles nearby (this requires very cold temperatures). If they interact, they entangle and previously fungible universes (being used for the computation) lose coherence with each other.

Entanglement is the loss of fungibility of two universes caused by a cascade of interactions in one that are immensely infeasible for to occur in the other.

connections_for_entanglement

Shor’s algorithm

An program to calculate the prime factors of a number efficiently on a quantum computer.

A quantum computer could factorise a number with more digits than there are particles in the universe (~3e80). The many small computations that happen in the algorithm are happening in separate very similar universes. At the end of the program these fungible universes interfere and combine (destructive interference) again and the result is obtained as the summary of all the separate computations.

A quantum computer that factors a number with more digits than particles in the universe will be a good demonstration that the computation is taking place by harnessing matter in a great many other universes.

connections_for_shors_algorithm

Universes differentiate all the time

Universes split all the time as particles move to possible new locations. More universes exist for the more likely new locations of the particles.

Each particle can occupy a number of positions that each have a probability given by the Schrodinger equation. Universes continue to differentiate to give rise to new universes with particles in these new positions. The measure of each position is not equal, with some positions more likely than others.

Universes continuously differentiate, with a greater measure of universes corresponding to the probabilities described in the wave equation.

connections_for_universes_differentiate_all_the_time

Only small particles can interfere

Rejoining of universes happens only to small things. It is not that a human could turn left and right in different universes and then rejoin. This is because there are too many particles in a human and they quickly do different things in each universe.

Large objects, such as people, have too many particles to remain fungible after differentiating into different universes. So the idea of a person taking one action in one universe and another action in another universe and then joining later does not occur.

Interference is only observed at small (particle-level) scales.

connections_for_only_small_particles_can_interfere

A particle behaves like an ink blot

Particles spread like ink. If you make the particle start off in a compact location, it will spread out faster initially. Then as it becomes more spread out, it will slow down.

Particles have a range of positions, speeds and directions (each with a probability as per the wave equation). A particle tends to spread out in space like ink. It it starts of in a small space it will rapidly expand, slowing as it spreads. So if you confine a particle to a precise location, you can expect the speeds and directions to increase, the more precise (confined) you make the particle. Each point in the ink blot corresponds to a universe where the particle is in that position. So while a particle can move anywhere, it tends to move in the region of the ink. A greater measure of universes exist where the particle is within the ink blob boundary.

Confinement of one property of a particle means that other properties increase in amplitude in other universes.

connections_for_a_particle_behaves_like_an_ink_blot

Diversity of attributes and Heisenberg’s uncertainty

There are always universes where the properties of a particle are different. Some universes will have different speed or positions. If you try to keep position precise, the speed of the particles in the different universes will be variable.

A particle will always have many universes where it’s attributers are different, such as location or speed. If you restrict a particle to a very specific location, there will be many universes where its speed will vary greatly. Practically, this shows up in experiments where the more you control one variable (position), the more diverse other variables become (speed). This can be seen in diffraction of light, where a narrower aperture causes light to spread more.

For a collection of fungible instances, some of their attributes must be diverse. For example a collection of fungible instances of a particle have several speeds. This means they will do different things an instant later. This is a consequence of quantum theory and is called Heisenberg’s Uncertainty Principle.

connections_for_diversity_of_attributes_and_heisenbergs_uncertainty

Motion is when the aggregate speed is not zero

Particles have ranges of speeds. A particle is moving if on average the speed is not zero.

When the aggregate value of the velocity of a particle is not zero, it moves. Velocity of a particle can also be considered as an ink blot similar to position. If the sum of all velocity is not zero, then across the multiverse, the particle is moving. Each universe has a slightly different velocity. Consider an electron fired in a certain direction. There will be some universes with faster, slower and angled trajectories.

Motion is when the aggregate velocity of a multiverse object is not zero.

connections_for_motion_is_when_the_aggregate_speed_is_not_zero

A simple atom

An electron forms a shell around a proton, where each universe has the electron at a difference place in the shell.

An electron existing as an ink blot can have a proton placed at the center of the field. The positive charge of the proton repels the electron and the ink blot changes to become an ink ‘shell’ surrounding the proton.

An electron occupies the space around a proton where the natural tendency of the electron to settle in a range of locations is balanced by the proton repelling it from all very central locations. The result is a hollow sphere.

connections_for_a_simple_atom

Multiverse enables quantum computer

The multiverse can be harnessed to outsource computations that all happen at the same time.

The histories of particles from other universes are used to perform the computation in a quantum computer. The flow of information is not confined to a single history. In Shor’s algorithm separate computations are initiated in an infinite different fungible universes. The fact that a particle has infinite possible universes that it can occupy provides a mechanism for vast computation to be performed in parallel.

Quantum parallelism is when different histories of computation combine.

connections_for_multiverse_enables_quantum_computer

Qubit

A particle can be set up to move two ways, one way in one universe and another way in another universe. The particle represents both these things happening and each movement is like a 0 or 1 in a normal computer.

Qubits have two properties:

Each qubit has a variable that can take one of two discrete values.
Special measures are taken to prevent the qubits from entanglement (cooling).

There are two universes, one for each state that can be held. Those universes can be combined after a computation using destructive interference. If many qubits were involved, the computation involves many universes.

Qubits are units that can encode a binary state. They can be composed with other qubits to represent all combinations of states. n qubits represents 2^n states.

connections_for_qubit

Quantum algorithm

Particles that exist in other universes can be combined with particles in our universe. Putting information on those particles can allow algorithms to do multiple things simultaneously.

The basis of the algorithm is to cause the information-carrying variables in some of the qubits to acquire both their values simultaneously. Consequently, regarding those qubits as a register, the number of separate instances of the register as a whole is large (for n qubits, registers is 2^n). Then for a period classic computations are formed, where waves of differentiation spread to some of the other qubits. Hence information is processed separately in each of the vast numbers of autonomous histories. Finally an interference process involving all the affected qubits combines the information in those histories into a single history. Because of the intervening computation which has processed the information, the final state is not the same as the initial one. The information is some function of the original information.

Connected information encoded units (qubits) can be caused to interfere in a controlled way. This allows the number of concurrent computations to rise exponentially (2^n) with the qubit count.

connections_for_quantum_algorithm

The problem with a rabbit forming while boiling tea

Anything can spontaneously happen, but a greater measure of universes are 'normal', where our explanations match reality.

Given that anything is possible, you could have a rabbit formed out of a pot of boiling water. As each particle happens to change in exactly the right way at the right time. This involves a large number of particles, and so there is a smaller measure of universes where this happens than where it happens for a mouse. In fact it is far more likely for a small number of particles in your mind to change such that you have a memory of a rabbit forming from the water, but the water did not change. So a hallucination occurs in a greater measure of universes. Greater still is the measure of universes where none of this happened and you simply made tea.

Spontaneous events occur in smaller measure of the universe. An explanation of those events involves a description of all the possible events also occurring in the multiverse.

connections_for_the_problem_with_a_rabbit_forming_while_boiling_tea

People become more alike over time

That people all correct errors means that over time they become more similar.

Knowledge grows along people. Knowledge is about error correction. There are more ways of being wrong than right. Knowledge creating entities rapidly become more alike in different histories. Aliens and AGI are likely to be curious about us, like we are about them.

People gradually become alike as they correct errors.

connections_for_people_become_more_alike_over_time

Chapter 12 - A physicists history of bad philosophy

The problem with postmodern philosophy

Postmodern philosophy is an attack on speaking clearly about philosophy.

Postmodern philosophy is characterised by the use of jargon in a way that seeks to obscure meaning. Terms are borrowed from other hard sciences and are used in ways that differ from their meanings used elsewhere.

Sokal published a parody paper in a postmodern journal to illuminate the problem with postmodern philosophy. It was published without being identified as a hoax.

The rise of the spread of bad ideas

There is an increase in the number of people rejecting the pursuit of knowledge. This is usually tied to their (mistaken) idea that knowledge comes from some authority.

Bad ideas are those that call into question the legitimacy of actual knowledge, science, the enlightenment and of reason itself. This is sometimes tied to the belief that knowledge that has arisen from an oppressing class is invalid. This is a misconception that works against the ability of society to correct errors and ignores that knowledge can be objective.

There is a modern movement to reject knowledge and the means to create new knowledge. This is tied to mistaken beliefs about how knowledge is created.

connections_for_the_rise_of_the_spread_of_bad_ideas

The problem with choosing science and rejecting philosophy

Philosophy is a topic that cannot be avoided. Any decision to not use philosophy is itself philosophy.

The idea that you can get everything you need from science and that philosophy has nothing to offer, is a claim that comes from outside of science. It is itself a philosophical claim. Hence philosophy cannot be avoided.

The rejection of philosophy is itself a philosophy.

Bad philosophy prevents people accepting multiverse

Thinking that prevents the development of new knowledge

Science currently focuses on the instruments and doesn’t seek to understand why.

TODO.

Doing philosophy

Doing philosophy is being aware of what you think and why.

Everyone has a philosophy (like I don’t need philosophy). Doing philosophy is bringing that philosophy that you have into your consciousness, and illuminating it so that you know what you think any given topic and why.

TODO.

connections_for_doing_philosophy

False philosophy is not bad philosophy

Bad philosophy that is not merely false, but which actively prevents the growth of other knowledge

False philosophical theories (e.g. empiricism, instrumentalism, kings have divines right, mathematicians can know certain truth).

Bad philosophy in some way prevents criticism and in some way prevents new knowledge from being created. Relativism is an example, because there is no need to find truth because there is no absolute truth (this is a problem because it implies there is no point to debate, because everyone has their own subjective truth). University philosophy can be bad philosophy because it directs people to original texts rather than latest editions of summaries of theories (this is a problem because it obscures truth and makes it hard to solve problems) and is more akin to history.

False philosophy is honestly seeking truth, but failing. Bad philosophy is dishonestly representing in a misleading way what you thought was truth.

connections_for_false_philosophy_is_not_bad_philosophy

People are flows of information that knowledge can grow along

People are something that can arise in the universe. They can do anything that is possible, and tend to do things they like. They cause things to happen, and the things they cause tend to continue to exist.

People are emergent quasi-autonomous flows of information upon which knowledge grows.

Quasi-autonomous (not completely free to disobey laws of physics and fly), because we can create knowledge about things that we are interested. There are a greater measure of universes where we do things that we like doing (because we have routines and preferences for them).

Flow of information because we affect things in a direction. Knowledge is that thing which tends to remain once it is there.

People are emergent quasi-autonomous flows of information upon which knowledge grows.

connections_for_people_are_flows_of_information_that_knowledge_can_grow_along

Photon experiments explain the multiverse

We did not expect the photon experiments to be linked to quantum theory. That they are linked is a good example of how good theories pop up in multiple places.

The experiment on photons explains the multiverse. It is a good example of reach, where quantum theory comes up in unexpected places not originally expected. Those experiments could be explained by other theories, but as the full quantum theory is hard to vary, they are consistent with the theory.

That the quantum photon experiments are explained by quantum theory is a demonstration that quantum theory has reach.

connections_for_photon_experiments_explain_the_multiverse

People who do not accept multiverse theory have bad philosophy

Quantum theory including the multiverse is not the majority opinion, possibly due to a culture where critical thought is discouraged.

Multiverse theory suffers from bad philosophy, where criticism and debate of ideas is discouraged. Common phrases in quantum theory classes are “There is no point asking”, “No one understands it”, “If you think you understand it, you don’t” or even “don’t ask why, you just need to solve the problems” instrumentalism. If this culture were different, perhaps more would accept it.

Bad philosophy may be hampering the acceptance of the full quantum theory of the multiverse.

connections_for_people_who_do_not_accept_multiverse_theory_have_bad_philosophy

Quantum theory came from two sources with different approaches

TODO

Two theories were developed that described atoms in new and surprising ways:

Heisenberg (uncertainty principle), describing the variables of particles.

Schrodinger (wave equation), describing the laws of motion.

TODO.

connections_for_quantum_theory_came_from_two_sources_with_different_approaches

Modern understanding of Heisenberg’s mistake

Heisenberg didn't know about multiple universes, so rather than seeing how he had a table of likelihoods of universes for different positions/speeds, he thought that observing a particle changed it's behaviour.

One of the two theories that led to quantum theory.

Idea: Physical variables of particles don’t have numerical values, they are matrices. The values in the matrices are probabilities that depend on your observations.

Modern understanding: The matrices are real, but each value represents the reality of a different universe in the multiverse.

Error: Observations you make don’t change the probabilities.

Heisenberg placed importance on the observer and their role in determining particle behaviour, rather than seeing how all his predicted variables were real but in different universes in the multiverse.

connections_for_modern_understanding_of_heisenbergs_mistake

Modern understanding of Schrodinger’s mistake

Schrodinger didn't know about multiple universes, so rather than seeing how he had an equation for knowing what measure of universes had particles with certain positions/speeds/interactions, he thought that he had an equation for the movement of particles as waves through space.

One of the two theories that led to quantum theory.

Idea: There is an equation for a particle that describes a wave moving through space. However for two particles there weren’t two waves as expected only one. Plus you could not split up that wave in to two or more waves. There was just one wave, in higher dimensional space.

Modern understanding: The wave describes what proportion of the instances of each particle (in each universe in the multiverse) are in each region of space and also the entanglement information of the particles.

Error: There is not a wave moving through space, rather if you were to view all the universes, together their aggregate action is a wave.

Schrodinger placed importance on particle being a wave rather than seeing that the wave was an aggregate view of the multiverse.

connections_for_modern_understanding_of_schrodingers_mistake

The rule of thumb that prevented multiverse from being pursued

A simple rule was discovered that let you predict what particle experiments would show. This may have caused people to avoid finding an explanation for why the rule worked. The "Shut up and calculate!" theory.

A rule of thumb was discovered that could be applied to schrodinger and heisenberg’s equations to get the same predictions and which matched experiments. “Whenever a measurement is made, all the histories but one cease to exist. The surviving one is chosen at random with the probability of each possible outcome being equal to measure of all the histories in which that outcome occurs.”. Rather than try to understand why this rule of thumb worked, this caused the scientific community to retreat into instrumentalism.

A parochial explanation for Schrodinger and Heisenberg’s findings contributed to the bad philosophy of instrumentalism seen among some physicists today.

connections_for_the_rule_of_thumb_that_prevented_multiverse_from_being_pursued

The problem with behaviourism in psychology

If someone behaves like they are happy or reports happiness, there is no way we know how happy they are.

Psychology often does not seek to find an explanation for mental states. Looking at reported happiness may just be studying other phenomena like “tendency to say they are happy, despite how they feel”. We do not know how to compare happiness between two people. The thing that is causing happiness needs to have an explanation in order to do good science.

Psychology is using instrumentalism when doing behaviourism.

The connection with happiness would involved comparing subjective interpretations, which we have no way of calibrating to a common standard. Not trying to find explanations for subjective things is instrumentalism.

connections_for_the_problem_with_behaviourism_in_psychology

The problem with scientific studies that have no explanations

A study looking at the correlations between things is bad science because it does not give an explanation for how things are happening in reality.

These studies are examples where the goal is to search for correlations that can be used to base decisions and understanding on, which is bad science because they lack explanations. They lead to the entrenchment of existing explanations, rather than being a process where new knowledge is created.

Consider a study of the genetic heritability of traits where one defines the heritability of a trait as its degree of statistical correlation with how genetically related the people are. According to this idea, one would have presumed in the past that being a slave was a heritable trait in America. This is a bad study because it finds correlations without explanations.

Or another study that shows that happiness is 90% heritable. However this has two problems. First, due to the problem with behaviourism, you cannot know if the people in the study are truly happy. Second, the genes could be causing any number of other factors that are not directly causing happiness. The genes found could be for facial bone shape that made people be called beautiful by others and receiving positive social interactions, thus leading to happiness. The consequence of the study might be to give up on trying to make people happy, however this would be false because you could change factors about how society perceives faces, and that could increase happiness.

Statistical correlations cannot inform what causes what, but one adds the inductivist equivocation that “they can be suggestive though”.

connections_for_the_problem_with_scientific_studies_that_have_no_explanations

We do not understand the qualia of animals

We do not know if animals suffer because we do not understand what it is like to be an animal or another person.

We have not yet created explanatory knowledge for qualia. So we cannot know the experience of animals. A study may look at if hunting animals causes them to suffer by using proxy measures, such as blood adrenaline, blood pressure and increased cortisol. It may conclude that these are seen in humans who report the qualia of suffering. However this is a bad scientific study because it does not have an explanation of how these measures are related to the qualia of suffering. These factors may be present in humans who are p;laying football and who say they are not suffering.

We do not have enough knowledge to decide if animals suffer.

connections_for_we_do_not_understand_the_qualia_of_animals

Chapter 13 - Choices

Actions in a government need to be accountable

It should be easy to point the blame clearly and cleanly if something bad is done by someone or some group in government

If something bad happens in government, the entity that made that decision or action should be easy to identify. Whether that is a person or an entire party, the ability to hold to account is important because action can be taken to remove that entity from the system. This corrects the error in the system (a bad actor).

Accountability allows for error attribution, which allows for error correction. This is important for a government.

Making the best decision is not about selecting the best option available

Decisions between options is about considering those options, but also trying to invent new options.

If there is a problem and some options are available, it is wrong to try and select between those options. There are always more options, because people create new ones. So there is never a perfect choice, for better choices may be created, in a creative and unpredictable way.

Selection of available options ignores that more options can be created at any moment.

connections_for_making_the_best_decision_is_not_about_selecting_the_best_option_available

For a group of voters, there is no system where the group can arrive at a fair result by voting. So, when groups try to make a decision together, they can only get 'pretty good' results, not 'perfect' results.

Arrow’s theorem shows that ideal social solutions results in logical inconsistencies. Decision making ideally involves using creating new solutions. The theorem is used to prove how choosing the correct seats per state for the House of Representatives in the United States leads to a strange conclusion: that there is no consistent way (you always have to round up or down in a way that goes against your original goals).

There is no rank-order electoral system that always results in fair choice of electoral seats.

There is no "will of the people". Applying maths to find the perfect formula for decision-making in society is not possible. There is no one special trick that lets all the preferences of society to be met.

Social choice theory is a branch of game theory that tries to apply mathematics to find the correct way to choose an option when people disagree. Once the best decision making process was found it could be set in stone. However it was found that this is not possible and that “there is no way to consider society as a decision maker with self-consistent preferences”.

Groups of humans cannot make perfectly rational choices, as illustrated by the apportionment paradox. The preferences of the group will always be inconsistent with other preferences of the group, as shown by Arrow’s theorem.

The problem with compromise

When the compromise fails, nobody takes the blame. Choosing a compromise option does not let the group learn if the original options were good or not. If the compromise fails, then the groups fall back to the original options and disagree again.

If two groups cannot agree between two options, then adding a third ‘compromise’ option creates a problem. If the third option fails, neither of the groups learns anything and both think that their original option is still the best option. Neither is accountable It is better to try (and refute) one of the two original options.

A compromise is bad because it prevents the identification, acceptance and correction of the error in the original options. The compromise is one that nobody believes in.

connections_for_the_problem_with_compromise

THe problem with weighing evidence

TODO

When there are two options, there are really two explanations (one for each option) that you are choosing. You examine the evidence for each explanation and then rule out options based on that evidence. You are then left with the best option available. You do not rate them and give them a score for weighing how good they are. Weighing leads to problems because large quantities of evidence for an older, more tested theory will outweigh the evidence for the newer, less tested theory. The older theory should instead be ruled out when compared to the better explanation and evidence for the newer better theory.

Weighing evidence when making a decision is equivalent to treating the mind like a population, where votes for different options are counted based on their evidence. This is problematic because Arrow’s Theorem shows how this can lead to illogical conclusions. Options instead are ruled out relative to other options.

Decision is about creating not weighing options

Making a decision involves creating the best decision, not weighing options

Decision making is about explaining evidence, not weighing evidence. It is a creative process where new conjectures are made to tie together other explanations and evidence. Those conjectures are criticised, and bad explanations are abandoned.

Decisions are made by conjectures about options, not the weighing of options.

connections_for_decision_is_about_creating_not_weighing_options

There is only one explanation for choosing best explanations

TODO

There is only one known good explanation for how to decide among good explanations.

TODO.

Groups of people cannot make perfect decisions, but they can remove bad decisions afterward. This can lead to progress, as bad ideas are repeatedly corrected.

Given that decisions cannot be perfect, the purpose of a system should be to always become better and better. The way this happens is to allow for bad decisions to be identified and corrected. Practically, this means it is not the goal to elect the correct political party/leader, but rather to be able to remove them if it is identified that there are problems with them.

Social systems cannot express a coherent opinion, but they can ensure that bad choices be fixed.

Two party systems are ideal

Two parties can iterate to find better policy over time. When they lose, they have nobody else to blame so they they can remove their bad policies easily and invent new policy openly.

Addition of more parties (3 or more total) encourages coalitions, which are a compromise that prevent accountability.
Proportional voting (and hence more parties) can lead to a failure of the voting to represent the view of the people.
Encourages parties to self-criticise when they lose because their policies must be the reason (rather than the policy of another in the coalition).
Any repeated losing encourages parties to create new policies and be very open about that.

Two party systems allow for very clear signals to the parties to improve over time. There is accountability when there are errors and clear incentives to change to become something that society wants.

connections_for_two_party_systems_are_ideal

Democracy is about peaceful dethroning

TODO

A good democracy can easily remove a bad ruler.

TODO.

The problem with proportional voting systems

With proportional voting, parties tend to be small and a bad ruler can be hard to remove.

Proportional voting systems have two problems. They:

Can cause the party with more voters can get fewer seats. This discourages large parties.
Can give disproportionate power to minority parties in a two-party dominated system. This incentivises coalition formation, which leads to compromise.

The formation of multiple small parties could lead to the situation where a party gets the most votes and wins, but only has a small percentage of votes (e.g., 10%). This party could be hard to remove as long as all the other parties get less than 10%. So while 90% of people think the party is a bad leader, they cannot remove them.

Proportional voting systems impair error correction.

connections_for_the_problem_with_proportional_voting_systems

Plurality voting is best

With plurality voting, parties tend to be big and a bad ruler can be removed the easiest.

Plurality voting, or “first past the post”, is where each district chooses one seat based on who has the most (“plural”) votes. It tends to give a disproportionately large vote count to the two largest parties. This reduces the likelihood of minority parties getting votes and leading to coalitions and compromises. While it can give the most votes to the party that had fewer total votes, this is less of a problem than the overall goal, which is to be able to remove a bad leader peacefully.

Plurality voting is better at peaceful dethroning than proportional voting because: it tends toward a two-party system, which allows unpopular parties to be voted out (if there are two, then the votes act to reject one over the other). The proportional voting system encourages multiple small parties, so there may be 20 parties, and the most popular might only have 10% of votes. So even though 90% of people don’t want them to win, they will be given power, and allowed to retain power if the situation keeps happening.

Plurality voting systems encourage error correction. Plurality voting prevents a situation where a bad small party cannot be removed. Voting for one seat per district prevents small parties from gaining and keeping power despite not having minority of votes.

connections_for_plurality_voting_is_best

Two party systems lead to policy convergence on objective truths

Two parties will tend to have similar policy the longer they improve their policies.

As parties try out and test new policies, they come to settle on ideas that have few errors. Both parties will end up converging on these good ideas, which is good because they will be present in policy despite who is elected. So while it may be tempting to be frustrated with parties because they seem too similar, another lens is to see that both parties have good policy that has been refined slowly over decades. Policies with fewer errors have more truth than ones with many errors. While it may be hard to pinpoint individual truths, the overall trend is “less error, more truth”. Policies move from misconception to ever-better misconception.

Parties in a two party system converge on the ideas that have the fewest errors. They both gradually improve to embody decades of testing and iteration.

connections_for_two_party_systems_lead_to_policy_convergence_on_objective_truths

Chapter 14 - Why are Flowers Beautiful?

Elegance is beauty of an explanation

Elegance is when an explanation for how something works is very pleasing to behold.

Elegance is when an explanation contains some deep truth and it is seen to be beautiful.

Elegance is the beauty of explanations.

Elegance is a heuristic for objective beauty

Theories that are elegant seem to be more likely to hold actual truths.

Elegant theories can be false or true. True theories can be elegant or ugly. However, it seems to be that if something is elegant, it is more likely to contain some fundamental truths. So beauty can be used as a guide in the search for truths. For example, many people describe mathematical proofs as being elegant. So finding elegant things in mathematics could help you in a quest for more proofs.

Elegance in science can be a useful guide to find truth.

The waste paper basket explains why art has objective truth

An artist discards things while they work. This is evidence that they work toward objective criteria.

The composer creates within the scope for some objective criteria within the scope of their discipline (music or painting). It seems to be the case that it is possible to be wrong.

That an artist discards intermediate works implies that they have an objective criteria.

Why theories with truth are elegant needs an explanation

Both scientists and artists use waste paper baskets and create beautiful work. If you can be wrong, then there is an objective. So because both can be wrong and both can create beauty, perhaps beauty is objective.

Scientists creating new theories have a waste paper basket that contains the intermediate works, much like an artist. That both creators (scientist or artist) work toward an objective criteria (both use a waste paper basket) and can produce beauty (beautiful art or elegant theories) suggests that beauty is an objective in both disciplines.

The analogy of artists and scientists in their creative endeavours points toward a common theme: that objective beauty exists.

connections_for_why_theories_with_truth_are_elegant_needs_an_explanation

Even savants may make mistakes

If someone doesn't appear to make mistakes, they may be discarding bad ideas in a way they find hard to explain.

While some scientists may make amazing contributions seemingly without making mistakes, it is likely that they are creating ideas and criticising them in their mind subconsciously. For example Ramanajan produced great theories without needing to make proofs, it is likely that he used conjecture and refutation at some level in his mind in a way that he could not communicate.

People who produce amazing works consistently and seemingly without making errors may be discarding ideas in their subconscious.

Some art is objectively better than others

It is clear that Mozart is objectively better than someone throwing rocks together.

Mozart is clearly better than stone age people banging rocks together in a rudimentary beat. It would be disingenuous to argue otherwise. If presented with two audio tracks for the rest of your life, which would you choose?

Some art is very clearly better at the extremes.

Some art is difficult to objectively compare, but not impossible

Deciding who is the best artist among modern artists may be hard for us, but in the future it may be doable.

It is hard to determine which modern artist is better than another. It may be possible in the future when we have better understanding of art. It is plausible that there is a very objective way to evaluate two different musicians using some very complicated system that we cannot yet imagine.

Difficulty in objective comparison between two art forms does not mean the comparison is impossible.

Attraction is evidence for objective beauty

If many people return a piece of art, it is a quality suggestive of objective nature of the art.

People are drawn to stay with a piece of art longer, or to return to it, or to travel great distances to experience it. This is a quality that some art has, which can be seen as a quality of good art.

Art that causes one to dwell is art that is more likely to be objective.

Creating new things involves creativity and criticism

Science and art both require creativity and criticism

Science without creativity does not result in new knowledge and art without criticism result in bad art.

Creating something new requires both creativity and criticism.

The mind can repurpose genetic mental predispositions

If there is a gene for some thought, it can be overcome.

If there is a gene for some mental trait, such as being afraid of heights, the mind can overcome that. This is because the mind is a universal computer, and so can find a new way of thinking that counteracts the effect of the gene.

There is no inherited thought that cannot be overridden by the universal nature of the mind.

Art is objective

TODO

You would prefer vivaldi to stones banging. A wastepaper basket of a musician shows they work toward something better. Subjective preference does not negate this.

TODO.

Preferences and objective standards both exist

TODO

We are yet to learn about all of the objective standards of art.

The mind can learn any new thought, so just how we can overcome any genetic predisposed thoughts, we can for thoughts about art. If there is a thought that something is not beautiful, it can be overwritten such that you consider it to be beautiful.

TODO.

connections_for_preferences_and_objective_standards_both_exist

Consensus is not a good measure for beauty

TODO

Not everyone agrees about what is the best scientific theory, even when there appears to be good reasons for people to agree. THis does not mean that there is no truth in the better theory, or that truth is subjective. Similarly, people may disagree about whether something is beautiful or not. Their disagreement does not mean that there is no objective beauty there.

TODO.

Flowers are a regularity in nature

There is something about all flowers we find appealing. There are not many ugly flowers.

We consistently perceive flowers to be beautiful, whereas this is not the case for most things. Not all rocks, or bodies of water, or animals are as consistent in our perception. We also do not find the roots and leaves of flower plants as beautiful.

Flowers are a regularity in nature.

The beauty of flowers are not cultural

We do not learn to appreciate flowers.

We find flowers appealing even if they are from places and cultures foreign to us.

Appreciation of flowers does not appear to come from cultural influence.

Flowers and flower-finding skills are hard to forge

To avoid forgery, both flowers and pollinators need to avoid imitations that lack the reward they need.

Flowers have evolved to be appealing to particular species for pollination, a slow and difficult process. If another plant could copy a flower, it would benefit from the pollinators it attracts. So flowers must be difficult to forge. Similarly, a pollinator evolves to be better at finding these flowers, and must be good at not being fooled by imitations with no nectar.

Flowers and pollinators both benefit if forgery is obvious to them.

Objective standards allow for forgery identification

If there is some clear criteria for something, it is easier to detect a copycat that is not quite right.

Flowers and pollinators both need to converge on systems that are hard to forge. If they converge on an objective system, this achieves that goal. So it may be that beauty is that objective standard they work towards. The plants and animals involved do not have a broad sense of beauty in the world, they just use beauty as a reliable way to prevent forgery of their genes.

Objective goals allow for reliable identification of imperfect imitation.

connections_for_objective_standards_allow_for_forgery_identification

People can discover objective beauty

Each plant discovers a single form of objective beauty through its flower. However people can see the beauty of all flowers.

People can understand the world and are able to identify things that have objective truth in them. So if flowers are objective standards, we would be able to identify the objective component across all flowers. Whereas the plants and animals involved are not aware of the beauty of all flowers, they have only discovered the objective beauty of the specific form they take.

Plants converge to an instance of beauty, people can perceive all the difference instances of beauty.

connections_for_people_can_discover_objective_beauty

Colour, contrast and symmetry are not beautiful

Unappealing things can have the same basic patterns that flowers have, so there seems to be more happening in flowers.

There are many very unappealing things that have colour, contrast and symmetry in nature (spiders). So it is not likely that these things are the reason flowers are beautiful.

Simple patterns are insufficient to explain the universal appeal of flowers over other things in nature.

Natural landscapes are a different sort of beauty

The appeal of a sunset is different from the beauty of flowers because if you change it, it still looks good.

Natural landscapes can be beautiful, but if you vary some component the beauty does not change. The waterfall does not become unappealing if you add or remove streams or rocks from it. Whereas modifying a flower can very obviously ruin the appeal.

Beauty in unevolved things is easy to vary, and is a different sort of beauty than flowers.

Perceiving objective beauty requires knowledge with reach

TODO

TODO make clearer

Perceiving the beauty of a mountain is subjective, but the beauty of a flower is objective. The knowledge that has been instantiated in the flower is information that tends to persist. FLowers tend to persist in a landscape, but a natural landscape can erode away and change. So telling the difference between the two kinds of beauty requires knowledge of the explanation for what objective beauty is.

TODO.

connections_for_perceiving_objective_beauty_requires_knowledge_with_reach

Art can have subjective and objective components

TODO

An image of a flower with a landscape is a mixture of objective and subjective components.

TODO.

Flowers are objectively beautiful

TODO

Flowers are an objective standard.

TODO.

Chapter 15 - The Evolution of Culture

Cultures are sets of ideas that cause similar behaviour

Culture is a group of ideas that make people more similar.

A culture is a set of ideas that cause their holders to become alike in some way. These ideas may be explicit or inexplicit (hard to put into words).

Culture is a set of ideas that causes conformity.

Memes are replicating units

TODO

Replicating unit of something, like genes or cultural ideas or practices.

TODO.

Ideas are imperfectly copied

Copying an idea is not a perfect process.

When an idea is replicated by moving from one person to another, the idea changes. The recipient creates a new idea based on what they think the original idea is. Being fallible, this is a process that leads to small changes, however if the idea is good, it will be hard to vary.

Meme replication leads to variations.

connections_for_ideas_are_imperfectly_copied

Some ideas cause behaviour

Some ideas cause you to do something, others do not.

You may have many ideas in a day, but only a few cause you to do something. If you receive a new idea from someone and it does not cause your behaviour to change, that idea is less likely to spread that another idea that caused you to take action and do something that helped spread the idea.

A subset of ideas you have influence your behaviour.

A culture is defined by a set of variants of ideas that cause behaviour

A culture is a collection of similar but slightly different ideas.

When people copy ideas, they are imperfectly copied, so a culture is defined by a set of variant that cause slightly different behaviours. Some variants make the recipient more likely to tell others, and others are very easy to copy without error.

A culture is the set of imperfectly replicated ideas that cause behaviours.

connections_for_a_culture_is_defined_by_a_set_of_variants_of_ideas_that_cause_behaviour

There are four types of cultural ideas

Ideas can range from fleeting thoughts to things that make you invent a new fashion.

Ideas with respect to their interaction with culture have four main types:

Ideas you have.
Ideas you have that cause you do something that nobody notices.
Ideas you have that cause you do something that others notice.
Ideas you have that cause you do something that others notice and then copy that behaviour.

The latter is an idea that goes viral and may persist for hundreds of years. Memes must cause behaviour, otherwise they will be lost.

Ideas can be private or they can drive their own replication though behaviours change.

The problem with arguments by analogy

TODO

TODO

Analogies require additional explanations to convey meaning.

Genes and memes are not similar

There is not a lot of useful truth in comparing memes with genes.

It is popular to compare memes with genes, however this is a bad argument because it is an analogy. The different parts of transmission, variation and selection are very different and there are no close analogues for the different parts of cellular biology to the more abstract memes.

Their only common element is the smallest level: they are both replicators that embody knowledge.

The popular analogy of genes and memes contains little truth.

connections_for_genes_and_memes_are_not_similar

Retold stories are examples of memes

Stories are remembered and retold, changing slightly each time.

A story is copied and retold in a way that introduces small changes. The story may evolve over time as different variants become more or less popular.

A story may evolve as it is replicated.

Creativity is evolutionary in nature in the mind

Smaller changes to ideas in the subconscious can lead to new creative ideas.

Blind variations and criticisms in the mind result in a creative thought at a higher level of thought in emergence.

Creativity may be explained by variation and selection of ideas in the subconscious mind.

Memes have two steps

Memes have two sources of error: Interpretation and replication.

Memes have two steps for a cycle of replication:

Person sees a meme and develops an idea about what is happening. This may be an imperfect interpretation of the original meme.
Person creates an action to enact the meme (speak, write or move). This may be an imperfect recreation of the idea of the meme.

Memes replicate when a person first develops an idea about what the meme is, then later does something to spread the idea to another person. Both steps are sources of variation.

Goal of society should be to become dynamic rather than static

Societies that encourage new ideas have a higher chance of solving harmful problems in the future.

Dynamic societies welcome criticism and new ideas. They can produce new ideas at a higher rate than static societies. This means that there will be a higher number of ideas to draw from to solve problems in the future. Those problems could be disastrous otherwise for society.

Dynamic societies develop technology faster as a result of the culture of criticism. Dynamic societies are stable under change over time.

The problem naturalistic preferences

Natural things are not necessarily good. Widespread memes are not necessarily good.

Some consider that natural things (plants, old diets) are better than artificial things (chemicals, synthetic things). This is a problem because spider venom is natural but harmful, and medicines can be artificial but helpful. This so applies to evolutionary things, where nature sometimes produces unhelpful or harmful things (appendix, or genes that lead to death or cancer). Similarly, memes can be harmful despite being widely distributed, such as ideas that harm others.

Natural or evolved things can be harmful. or not helpful.

Biological evolution is just a preface for memetic evolution

Evolution of ideas is likely to have happened more than it has in physical organisms.

Memes evolve inside the mind of people during the process of creativity. They also evolve between people. The majority of all evolution has therefore occurred not in biology, but in the minds of people, who can create idea variations and select them hundreds of times before sharing the idea with others. Hence, the amount of evolution that is still to take place in the minds of people is vast.

Minds, not biology, is where evolution has happened the most.

Some memes can be intelligently designed

Some memes can be carefully designed.

Some memes can be carefully crafted specifically to be easy to spread and replicate.

Memes can be intelligently designed.

The problem with static societies

Static societies prevent new ideas by having systems to prevent people from being creative.

Static societies have customs, laws and taboos. They prevent new ideas and discourage changes to existing customs. The way they do this is by preventing people from being creative. This usually involves teaching children things that prevents them from being creative. This can be through systems that lead to obedience, piety or duty. People work to express their creativity to more faithfully and accurately enact existing customs. For example, taking an existing behaviour and perfecting it. If everyone does this, no new ideas are created.

Static societies disable the ability of people to be creative.

Dynamic societies are characterised by respectful criticism

Societies that allow change do so by welcoming calm criticism of ideas.

Respectful criticism is the mechanism that dynamic societies use to introduce new ideas.

Respectful criticism allows for dynamic change under stability.

How dynamic societies remain dynamic is poorly understood

There are things in our society that help it to be stable despite so many new changes.

For a dynamic society to continually allow new ideas to arise and for criticism to occur, there must be some constant feature that allows this. This stable thing allows the new and dynamic ideas to come without disrupting the society and causing it to revert back to a static society. Culture (norms) and cultural institutions (universities, governments) are long-lived features in society that can help to maintain the culture of criticism. What these factors are exactly is not fully understood.

Dynamic societies remain stable under change due to knowledge that exists implicitly in structures within society. This knowledge is not well known.

Useful memes tend to survive

People adopt memes if they are useful to themselves.

Memes are likely to take hold if they are interesting, funny, elegant, easy to remember or morally right in the eyes of people, who have diverse and unpredictable objectives. So memes that meet this criteria are useful to those people. Memes will survive if they are truths with reach. Ideas with reach are remembered and repeatedly used by people in their lives. Other competing ideas are tried and all ideas are criticised freely and openly. The rival bad ideas that do not survive criticism will not be replicated as much as useful ideas that do survive criticism.

A meme that is appealing to a diverse group of people is useful.

connections_for_useful_memes_tend_to_survive

Rational memes are produced by rational processes

TODO

Rational memes are those ideas created and replicated using rational and critical thought. These are present in dynamic societies. In the absence of criticism, these ideas lose their advantage over other less rational ideas and can become lost.

TODO.

Antirational memes

TODO

Antirational memes are ideas which disable the capacity of their holders to criticise themselves. This prevents progress and they are present in static societies.

A persistent idea that prevents people to criticise ideas. Society that encourages static culture, where creativity is used to adhere to standards.

TODO.

connections_for_antirational_memes

Memes must match the society to survive

Free thinking society shuts down bad ideas. Closed society cannot promote good ideas.

Antirational memes do not survive in rational society (they are criticised and are shown to be worse than other ideas).

Rational memes do not survive in antirational society (they are not criticised and shown to be better than other ideas).

Rational societies become filled with more rational memes and vice versa.

Static cultures cannot become static after periods of being dynamic

Once a society temporarily adopts rational memes, it cannot become static again.

TODO Improve.

If North korea were to adopt rational memes, it would be unable to return to its current static state.

TODO.

Antirational memes can be detected by the shame they cause

Society has outdated beliefs and customs that don't make sense, but if you don't follow them you feel shame. Even if you know about the meme.

Rather than create their own rituals, behaviours and goals, society has collections of antirational memes that have been passed down. These memes are antirational because they cannot be challenged (it is socially discouraged implicitly or explicitly). The pressure to comply with these memes can manifest in a way that the person feels shame or embarrassment if they do not obey the meme.

Challenging antirational memes often leads to the person feeling shame.

Transition from static to dynamic society happens slowly

Moving to an open society can be slow as existing ideas and behaviours stop the new ideas from flourishing with open criticism.

Transitions are complex and rational and antirational memes compete and interfere with each other. The different memes cause behaviour which disrupts the accurate replication of the other sort of meme. Thus a transition is a period where memes have high variability and may take time to correct.

Moving from a static to a dynamic society involves introducing new memes and behaviours which disrupt meme replication.

The Enlightenment is when the world is mainly transformed by explanatory knowledge

The Enlightenment is when the world changes mostly due to our ability to create explanations for how things work, rather than other processes.

The Enlightenment is a time when most things are being transformed as a consequence of explanatory knowledge.

The Enlightenment is the moment at which explanatory knowledge is beginning to assume its soon to be normal role as the most important determinant of physical events.

The problem with the collective inventor

Inventions happen by evolution of ideas in a single mind, not across multiple people.

Most inventions are the result of a single person engaged in creativity. A person takes existing ideas and creates a new idea in their mind using creativity. Thus even is there are multiple people working on a project, small ideas are individual inventions by individual people who create them. This contrasts with the idea that ideas are all group efforts.

Inventions have sole inventors, not groups of inventors.

connections_for_the_problem_with_the_collective_inventor

The problem with wokeism

TODO

TODO - not sure where this part is in the podcast.

Wokeism prevents criticism.

TODO.

Chapter 16 - The Evolution of Creativity

People are cosmically significant

If you want to know what happens to the sun in the future, you need to know about people, because they could do things to it that physics doesn't predict.

The laws of physics are not enough to forecast things in the universe, we must consider people. The fact that people could come along and create new unpredictable knowledge and use that to transform the physical world means that the future is unpredictable.

People can have enormous impacts in the future. These cannot be predicted.

Creativity is the only process that can produce knowledge

Creativity is the process that gives us knowledge.

Philosophy, mathematics, art and science all come out of creativity. It is how we create explanatory knowledge. Animals do not have creativity. Creativity allows us to create a model of anything in the universe, and have that model increase in accuracy over time.

Creativity is the core skill that people have. We use it to create explanatory knowledge.

connections_for_creativity_is_the_only_process_that_can_produce_knowledge

We don’t yet understand creativity

If we understood creativity we could write a program to be creative.

We do not fully understand creativity. If we could, then we could program it. This is similar to not knowing how to program evolution of minds in general.

If you cannot program it, you do not understand it.

connections_for_we_dont_yet_understand_creativity

Human choice is a product of creativity

People can make choices by adding new unpredictable options. Thus free will and choice arise from creativity.

If there are choices to make in the universe, people are the entities that can choose them. Many people have the sensation of free will and it may be that free will is possible because we create knowledge. The mechanism for this is potentially that when there is a choice to be made, people can generate new options that were previously not there. This is a process of creativity, which we don’t fully understand.

Creating new choices is equivalent to free will.

connections_for_human_choice_is_a_product_of_creativity

Consciousness, free will and creativity may be interconnected

People create knowledge about how things work, unlike all other animals. We also have consciousness, free will and creativity sensations, and perhaps they are all related.

People have behaviour of creating new explanatory knowledge. This is evidenced by the vast amount of technology and society around us, which is the beginning of infinity in terms of potential. We have this ability and we also seem to have very clearly obvious phenomena that appear to have come along with the “jump to infinity”. So consciousness, free will and creativity may be interrelated phenomena that are different aspects of the same fundamental skill we uniquely have.

Creating explanatory knowledge, consciousness, free will and creativity may all be phenomena arising from a single skill that people have.

Humans are saviours not destroyers

Humans help other animals and only humans can hope to prevent future disasters.

We are the only species that consistently helps other species. Without humans, the Earth will experience any number of terrible and harmful events for life on Earth. We have some small side effects on some animals and habitats, however, those species will eventually be threatened at some point in the future anyway. It is our ability to create explanatory knowledge at high enough rates that can create solutions to save the world.

If there is going to be a threat to life on Earth, no other species can save it.

connections_for_humans_are_saviours_not_destroyers

The emergent level explanations are real explanations

The actual explanation of something is the emergent one.

The explanation as to why, at the higher emergent level, is really the explanation. It is not that some things were determined happen because the big bang happened and ever since there has been laws of motion acting. That sort of deterministic explanation for the copper atom being on the nose of a statue is not an explanation. It is a general purpose explanation for anything in the universe. The better explanation is the high level descriptions of emergent things that are happening. So the better explanation involves the story of people, history, explorers, monuments and statue construction materials. These things are all real, and so this is the real explanation.

Explanations at the emergent level are real explanations.

connections_for_the_emergent_level_explanations_are_real_explanations

Predicting the emergent future in a deterministic world is not possible

You cannot predict what complex systems will do as a whole, so unpredictable will is free will.

The universe has deterministic laws, but the emergent level is the real explanation for what is happening. There is no thing that can predict what the future of the mind is, for a simulation of that thing would require that you know what all the particles are doing at the individual level. This is a problem because you do not know which subjective universe you end up in at each branch in the multiverse. The future you end up in is random from the perspective of our universe. No such system exists that can predict the emergent behaviour of the universe. So the agents in our universe are really unpredictable, and their choices cannot be predicted. In this way they are free to choose.

Practical unpredictability really is freedom from prediction, or free will.

connections_for_predicting_the_emergent_future_in_a_deterministic_world_is_not_possible

Creativity evolved only once

Creativity evolved only once.

There is no other example of creativity in nature, therefore it is not a convergent feature of evolution.

Creativity appears to have evolved once.

Humans failed to progress for thousands of years due to static society

Human ancestors were creative but tried to perfect existing culture for social reward rather than inventing new things.

Creativity evolved in human ancestors but it does not appear that it was used to invent technology. Creativity was likely being channelled to be the best at a current meme in society. Early society was likely static, where new ideas are not rewarded and new inventions were taboo. Perhaps innovators were cast out of society.

The absence of evidence for innovation in primitive human society can be explained by it being a static society. Social reward was given to those who could enact existing customs more faithfully than the norm or display exceptional obedience to existing ideas. Thus creativity was used in selection of mates and survived, but no significant innovation occurred.

connections_for_humans_failed_to_progress_for_thousands_of_years_due_to_static_society

Animals can only copy

Animals have systems that allow them to observe and copy behaviour. The systems are inherited and they only work for a narrow range of behaviour, unlike humans, who have a general purpose system in our mind.

Animals have a large repertoire of actions they can use to mimic. They do not form ideas about what is happening. Humans are universal explainers. The things that they can understand are limited to what their genes permit them to understand. We have no limit.

Animals have inexplicit knowledge (from genes) that allow them to understand how to replicate behaviour that they observe. They do not create knowledge during the copying process.

Meme acquisition comes from guessing

Memes are guessed and attempted, not copied.

Meme acquisition is not imitation. It is error-prone guesswork.

Observation of a meme.
A guess of what the important component is.
The meme exists in the brain.
Action to enact the behaviour of the meme.

The guesswork component involves creating knowledge because you form a new idea about what is happening. Humans observe behaviour and then form an idea about what is happening. Then they attempt to enact this behaviour according to that idea. This is unlike a monkey which simply imitates the movement.

Memes are adopted through a creative process in the recipient.

connections_for_meme_acquisition_comes_from_guessing

Animals learning is explained by behaviour passing

Animals have inbuilt libraries of actions that can be combined to form new actions. This is a slow and limited process.

An ape can copy another ape with a better way to crack a nut. This is an example of behaviour passing. This is not conjecture and refutation. There is a repertoire or library of behaviours that can be combined in new ways not previously seen in a group. This requires lengthy watching and attempting and is an inefficient process. Humans do something completely different, we seek the meaning of what is being done, not the specific movements or sounds. So we might watch a lecture and remember the concept but not a single exact sentence from it.

Animals can use recombination of basic inherited behaviours to mimic or produce previously unseen behaviour. A process different from humans in that it doesn’t involve creativity.

People that behave the same have rediscovered the same idea

If you behave like another person it is because you have created your own interpretation of what they are doing, and come up with a behaviour that you think represents that idea. You did not copy, you rediscovered.

Using our explanations we see through the behaviour to understand its meaning. Parrots copy sound, apes copy movement but humans do not copy behaviour. They use conjecture, criticism and experimentation to get at the meaning of things (others behaviour, their own, and the world in general). This is what creativity does. If we end up behaving the same as other people it is because we have rediscovered the same idea.

Behaviour replication in humans is rediscovery of the idea behind the behaviour.

Creativity was good for memes and enabled a second benefit, to create knowledge

Creativity allows us to get existing knowledge (memes around us), however a side effect is the ability to create new knowledge.

There are two puzzles with one solution (creativity):

Why human creativity was advantageous at a time when there was almost no innovation.
How can memes be replicated given they have content that the recipient never observes.

What replicates human memes is creativity, and creativity was used while it was evolving, to replicate memes. Creativity is used to acquire existing knowledge, not create new knowledge. But the mechanism to do both things is the same. So in acquiring the ability to do the former, we automatically gain the ability to do the latter. This is an example of reach, and enabled everything that is uniquely human.

Creativity arose because it allowed meme replication. A by-product was creating new knowledge.

Both science and meme understanding both require discovery of a hidden explanation

Discovering a hidden explanation is important both in understanding memes and in doing science.

The same process to understand a meme is used to do science. Discovering a hidden explanation.

Meme understanding. Discovering an explanation for the ideas in the mind of other people.
Doing science. Discovering an explanation for a regularity or law of nature.

Both have access to evidence, with which explanations can be tested (behaviours for memes and physical phenomena for science).

Discovering hidden explanations is central to memes and science.

Ancient toolmaking is likely to be ritual based

Rituals are a way for strict societies to improve a set of useful tools.

Memes in primitive human society would likely have actual benefit for survival. For example, making a spear might be a ritual that people would try to do to be part of the tribe. You could apply creativity to make an extra good (strong, light or sharp) spear. The outcome might be a tribe that is very strict about spear-making and consequently very good at hunting and fighting. So strict rituals for this static society would lead to better chance at surviving and growing, despite discouraging new ideas.

A static society can excel at narrow tasks through rituals that harness creativity.

Success in a static society is for those who replicate memes the best

A primitive society would reward and select for more creative people.

Those who can replicate the memes of a static society would impress leaders, peers and mates in that society. So those who could do that (using creativity) would eat, thrive and reproduce more, thus ensuring creativity was passed on.

Creativity can a survival advantage in a static society. So even societies that discourage innovation select for people with increased ability to innovate (they can replicate memes better).

connections_for_success_in_a_static_society_is_for_those_who_replicate_memes_the_best

There cannot be a final theory of physics because there is implicit knowledge in the form of any idea

For any final theory, there are more questions about that theory.

Even if a final theory was successful there would always be more to ask. “Why does the theory have the form that it does?”.

No final theory can answer all questions about that theory. Every new conjecture has explicit and implicit knowledge.

Chapter 17 - Unsustainable

New resources are created using knowledge

Anything around us may at some point be very useful to us, though we cannot know what or how in advance.

If we invent a new technology, some thing may go from being non-resource to resource. Uranium was previously just a mineral in a rock to us, but now it can be harnessed to generate electricity. It is thus now a resource where it previously was not. Given we cannot predict knowledge creation, we do not know what things around us can be resources.

New resources are defined using knowledge.

connections_for_new_resources_are_created_using_knowledge

Energy sources are transition energy sources

Our energy sources may all be replaced by something better when we invent it.

As we create new resources, we may find new sources of energy. Thus any energy source we currently know about may be a transition source before the new onw.

Energy sources are all transition sources.

connections_for_energy_sources_are_transition_energy_sources

The problem with Easter Island comparisons

Easter island failed because it discouraged new ideas and failed to recognise bad ideas. It is not a good forecast for what will happen to Earth because we can encourage new ideas.

Easter islanders appeared to have been cutting down their last trees to make roads to move large monuments around when they went extinct. Had they stopped and challenged this practice, they may have survived. Their civilisation failed as a result of being a static society unable to criticise their bad ideas. If they were a dynamic society, they may have been able to create better ideas (that lead to surviving) rather than making monuments and monument infrastructure for the Gods.

Easter Island is a static society which was unstable when faced with new problems. It is a variant of Spaceship Earth. The environment is not a welcoming place, but we can survive if we create knowledge to solve problems.

connections_for_the_problem_with_easter_island_comparisons

Human sustainability is about constant and rapid progress

Sustainability is about doing things that provide energy and resources for civilisation. It is not about maintaining status quo.

Sustainability is not about sustaining things in a static way. It is about having the ability to provide us what we need to sustain. So if there is something that we need, we should prioritise that, even if it means changing the world around us. We need to make fast progress if we want to solve more problems. This requires energy and resources.

Sustainability is about the ability to meet real needs. We have the ability to solve any problem and will need an ever increasing collection of knowledge and resources to achieve this.

connections_for_human_sustainability_is_about_constant_and_rapid_progress

Static societies can cause its members to suffer

Societies that don't welcome ideas can become stuck in behaviours that cause harm to its people.

Some societies may be stuck with culture that causes its members harm. Easter Islanders, valued stone monuments above forests and ultimately people. They were unable to create new ideas to reduce or prevent their suffering. It is unfortunate that this culture lasted so long, for if they had a dynamic society they could have solved their problems and been around today.

A static society can be unable to rapidly solve problems to save itself and prevent harm and suffering.

connections_for_static_societies_can_cause_its_members_to_suffer

The problem with trying to learn morality from all cultures

It is unlikely that a society closed to new ideas can be studied to learn new ideas. This includes ideas about what we should focus our attention on and do in life.

Progress is sustainable indefinitely, but only be people with a culture of criticism in a dynamic society. This is an optimistic society who hope to learn from their own failings. However, learning from the failings of Easter island is not likely to be valuable for they made no progress in their pursuits. We should not hope to learn about new technologies or advancements from them. Similarly, other cultures that prevent criticism (such as those who mutilate and subdue some members) are unlikely to have accumulated advanced knowledge because they fundamentally have a static society that resists new ideas.

A static society is unlikely to have developed advanced knowledge in any domain. Progress requires the openness to new ideas. So to value all cultures as being equal (moral relativism) incorrectly expects all cultures to have made moral progress.

connections_for_the_problem_with_trying_to_learn_morality_from_all_cultures

Ideas are abstract yet have real impact on the world

Ideas are not physical things because they can exist in writing or in sound, but they can affect physical things (a bridge can happen because of an idea).

An idea exists as a crackle of electricity in neurons in the mind. But it is not identical to that. For you can take the idea and write it down on a piece of paper (pencil on paper), or you could explain it (sound waves). The idea is an abstraction yet has physical effects on the real world (ideas can cause people to do things).

Ideas are real and abstract and can be instantiated in different physical forms.

Societies improve slowly

Society can become better by trying out new ideas and seeing how they go, undoing if need be.

Societies improve incrementally. Every step is hopefully for the good but may be bad. If a bad change is put in place then it can quickly be undone.

Society improves incrementally where new ideas are created and tested.

The problem with the Great Reset

Sudden change in society is undesirable because it gets rid of good things. It is better to slowly evolve changing what you think is broken.

People want to get rid of institutions and to redistribute wealth. The problem with this is that what we currently have (wealth, health and success) has come from these institutions. So they should not be discarded, but incrementally improved. Moving too fast could destroy things that keep our society stable under change, because these stabilising factors are poorly understood.

Sudden revolutions get rid of good and bad things. It is better to change things in a gradual way so that you can see the effect of your actions and undo them if need.

connections_for_the_problem_with_the_great_reset

Institutions may have harmful and beneficial components

If there is some part of society that seems to be causing problems, it should be slowly changed because it might also be doing lots of good. If you remove it quickly it might cause severe problems.

Any institution in society (organisation, structure or formal convention) may have harmful components. It may discriminate against groups, or cause harm or problems somewhere. However, that same institution may be causing benefits, such as stability of important structures for education, criticism or preventing violence between groups. If that institution is removed, then it is difficult to predict if the effect will be net positive or negative. Hence institutions should be subject to evolution, where new ideas are proposed, criticised and tested, with bad ideas being reverted.

Institutions are complex and may have critical benefits as well as causing harms. They should be evolved critically rather than removed suddenly.

The problem with antihuman education

By teaching that people are bad and only cause harm, you may prevent people from believing they can do good things. If nobody tries to make great things, nobody will.

Teaching students how humans are the cause of many problems is a pessimistic viewpoint (we are in the Holocene, we are a genocidal species, we destroy the environment). There are many great and important things that humans have done and can do in the future. By only presenting pessimistic views, fewer people may be excited to try and create those positive things.

Exclusive antihuman pessimism in education may dissuade people from trying to create positive impacts.

The problem with the crowded planet argument

We can have more people on Earth and have even better living standards in the future. This is because we can always solve new problems and create new ideas for new resources.

People argue that the planet is crowded and that “the resources are going to run out”. They have real scientific evidence for specific resources.

However, because we are going to solve those problems and create new resources that means we are going to be able to both sustain more people and increase standards of living.

That resources are defined by our knowledge, which is unbounded, is a good reason to believe the carrying capacity and living standards for the Earth is much higher than what we have today.

connections_for_the_problem_with_the_crowded_planet_argument

It is okay that we do not know how to solve some problems

Don't worry if there is a problem that doesn't seem solvable. It is possible that it will be solved, and you cannot predict that.

It may be concerning to some that we have problems today without solutions. However, the scope of new things that can be invented in a day, week or year are completely unknown. Right now there could be someone creating a new piece of knowledge that has an effect so dramatic that the problem you are thinking of is no longer relevant. To worry that the problem is not possible to be solved in time is wrong, because it is possible. To list reasons why it is not likely is to boldly ignore the list of possible new ideas people may have that could solve or change the problem. That list of possible solutions awaiting invention is not predictable, and worrying about unknowable things is a fruitless endeavour.

It is possible that any currently unsolved problem will be solved in time.

connections_for_it_is_okay_that_we_do_not_know_how_to_solve_some_problems

The problem with doom prophets

TODO

It involves predicting the creation of knowledge which is impossible.

TODO.

The parable of Europium

Resources that seem to be running out may not be important. We don't know that the scarce resources will be needed in the future.

It was thought that televisions were going to deplete the element Europium, causing terrible social problems. However different televisions were invented and the element was no longer needed. Hence Europium did not cause a class divide of television and no-television people as expected.

It cannot be predicted if a resource will be scarce in the future. Unpredictable inventions may reduce the need for any given scarce resource, as was the case with Europium used in televisions.

New technology should not be feared as a cause for class divide. Innovations can make it available to many people very quickly.

It is common to predict that there will be two classes of people in the future, the Have’s and Have-not’s:

Advanced medical technology
AI as an assistant
Life prolonging technology
Genetic designer babies

However this is refuted by the parable of Europium. Which shows that eventually new technology becomes democratised after being funded by the initial selling of expensive and limited availability technology.

Technology is usually democratised after initial funding and subsequent innovation and democratisation.

connections_for_the_problem_with_predicting_social_inequality_based_on_technology

The problem with a pessimistic view of people

People are the solution to the problems of tomorrow. While every new solution comes with a new problem, we can solve that problem too. Each solution brings into a better world.

The pessimistic view of people is that they are wasters who take precious resources and madly convert them in to useless things (like Europium into coloured televisions). This is true of static societies, where un-criticised ideas can lead to harmful actions (like the stone statues and tree-roads on Easter island).

The optimistic view of people is that people are problem solvers. They are creators of the ‘unsustainable solution’ and hence of the solution to the next problem.

In the pessimistic option, problem solving is a disease and sustainability is the cure.

In the optimistic option, sustainability is the disease and people are the cure.

The pessimistic view of people views any solution we create negatively, because it brings new problems. However, this ignores the initial solution that brought us forward from lesser past. It ignores that we make gradual progress and encounter new problems that we can always address.

connections_for_the_problem_with_a_pessimistic_view_of_people

Progress is not inevitable, it requires hard work

We have to work hard to solve problems to solve them in time.

We will encounter problems and they will be soluble, but that does not mean that they will be solved, or in in time. They require hard work, resources, effort and the right culture.

The solubility of problems does not guarantee a timely solution will be found.

connections_for_progress_is_not_inevitable_it_requires_hard_work

The future cannot be scientifically planned

We cannot use science to plan the future. That assumes we know what new knowledge will be created.

We can make scientific projections, but these require that we make bold claims about what future technology we will have in the future. The further out we plan, the more likely it is that some knowledge will be created that will affect that plan in some way.

Planning the future is equivalent to prophesying the content of future knowledge creation.

The problem with economic predictions from climate change

We cannot predict the economic impact of climate change because anything we invent could affect those predictions.

Projecting economic impacts of climate change is not possible because any innovations we have that affect the situation will have an effect. So the projections will depend on prophesying what knowledge will be created.

Economic forecasts of climate change require the forecasting of new knowledge that affects the situation.

Technology to cool the globe should be the focus of climate efforts

TODO

Slow the rise of CO2 levels may help but solutions are required.

We need to exert effort to solve the hard problem of cooling the globe. Slowing the rise of CO2 is insufficient to solve the problem.

TODO.

Increasing energy consumption can increase knowledge and therefore resources

TODO

People create knowledge to define resources. By using more energy to create knowledge, we can create new resources.

TODO.

Reducing consumption alone is insufficient to save environment

TODO

TODO

TODO.

Increasing wealth is better than reducing footprint

TODO

Giving computers to all humans is better even if it produces carbon.

TODO.

Chapter 18 - The Beginning

The infinite journey begins with a choice to begin

We can set out on an infinite journey if we choose.

Earth is not the eternal and only home of mankind. It is a starting point of an infinite adventure. All you need to do is make the decision to static society. That would be the end of eternity and the beginning of infinity.

(Paraphrase of Asimov end of eternity 1985).

Choice to make society dynamic makes it possible to start progressing toward infinite knowledge.

Problems we do know about are less risk than ones we don’t know about

Problems we can see are may be less dangerous than problems we haven't thought of yet.

There are many problems we are aware of and many people focus on trying to solve these problems. To some extent this makes those problems less likely to cause harm, because creation of solutions is being attempted. Problems that nobody is trying to solve pose a greater risk, and this includes all the problems we are not aware of.

Attention on problems acts to partially mitigate their potential harm.

The problem with hysterical reactions to problems

Making exaggerated negative claims about problems makes people afraid and turns people off listening to scientific ideas in the future.

There are two problems with reacting hysterically to problems. First the reactions frighten people. Second they discourage people from listening to scientific theories in the future (when claims do not come to pass, it is like the “boy who cried wolf”).

Hyperbolic claims about problems causes unproductive fear and actively discourages engagement with future scientific theories.

A culture of criticism requires training people

TODO

There is no trend where things automatically keep getting better. People must be able to think of new ideas because they are free to do so. They have been trained to think that it is okay to think of new ideas.

TODO.

We have infinite theoretical knowledge to discover, just like the physical universe

The potential knowledge we can discover is as infinite as the things we can discover in the physical universe.

It is common to believe that there are many more things to learn about the Earth and the physical universe, despite already knowing a lot. However people have not yet applied this thinking to theoretical knowledge. There is an infinite amount more knowledge that we have not yet discovered. The theories we currently have are not “nearly there” or “halfway there”, there are infinite more theories to create.

It is uncommon to acknowledge that there is an infinite amount of theoretical knowledge yet undiscovered, yet common to see that the physical universe has infinite to explore.

Each person has an infinite amount they do not know

All humans share that we have only a miniscule amount of knowledge that is possible.

Everyone has a small slice of specialised knowledge, however this is tiny compared with the infinite amount of knowledge we do not know. So all humans, including the most famous thinkers, have this in common. We are all at the beginning of infinity in this regard.

Each person has some knowledge that is closer to the start of learning than it is to infinite knowledge potential.

Principles can be both emergent and fundamental

Some important things (evolution or people) arise from complex systems and are important in many other fields.

The process of evolution by natural selection is critical within the domain of biology. It arose in an emergent fashion from the more fundamental laws of physics and is also fundamental (has effects on many other domains). People are also emergent and fundamental in a similar way.

Some principles, such as Neo-darwinism or the phenomenon of people, are both emergent and fundamental.

connections_for_principles_can_be_both_emergent_and_fundamental

The problem with plans for utopias

There is no recipe for a perfect society because it changes unpredictably. So a recipe would need to constantly change to meet changing needs of the people.

A finite number of well known changes can bring about a perfected human state. This leads to dogmatism and tyranny because there are no final human needs. There will always be new problems and needs to solve for. So any plan cannot deliver these. All that can be done is to work with a plan to address problems as they arise.

“From each according to his ability to each according to his needs” (Marx) is problematic because the needs of people will constantly change as people create new ideas and want different things.

Utopian planning fails to account for the needs of humans changing.

Creativity requires optimism

If you are pessimistic, you are very unlikely to be creative.

Almost no one is creative in fields in which they are pessimistic. If you want to make progress in a field, a good first step is to be optimistic and believe that you can make progress. This may seem trivial, but in the extreme it is apparent that if you believe strongly that no progress can be made in an area, you will not be open to even consider new ideas.

Optimism is often a precursor to new creations.

We grow our knowledge growth by problem hopping

We grow our knowledge by moving from old problems to new and better problems.

The growth of knowledge is a continual transition from problems to better problems (not solutions). This is an extension of the idea that all our theories can be thought of as misconceptions awaiting improvement. It can be thought of as a challenge to improve things (solve the problem of what the error is in any given theory). People are ideally open to change their minds on things rather than believing scientific dogma.

Knowledge is gained by focusing on an infinite series of problems.

connections_for_we_grow_our_knowledge_growth_by_problem_hopping

All theories are not equal

We don't know how much truth is in each new and better theory we create. All we can do is say that we make progress.

All theories are not equal. We move from misconception to ever-better misconception, however we cannot say how much truth we have gained, or what fraction of a theory is truth. All we can do is correct errors.

We cannot quantify the truth in our theories.

connections_for_all_theories_are_not_equal

Fallibilism does not require passivism

Believing that your ideas can all be wrong does not mean passively letting them go. It is ok to claim that there are ideas which are worth protecting into the future.

Being a fallibilist does not mean that we should be passive when faced with other peoples strong opinions. It is appropriate to be passionate about ideas we have strong convictions about. We can claim to know what we know and defend moral ideals and values that we think are important for the maintenance of civilisation.

Fallibilists can have conviction and be more active than passive relativists and tyrannical dogmatists.

connections_for_fallibilism_does_not_require_passivism

Objective truths exist in all disciplines and finality in none of them

All fields (science, art, literature, etc.) have some truths. However, any idea may contain partial error so ultimate truths are never possible. It is also important to be wary of ideas that are vague or contradict themselves.

All disciplines have objective truth. This does not mean that there are conclusive and final things that can be reached in any of them. Science cannot have final theories and proofs, and neither can there be final theories in literature. People think that non-scientific disciplines have no truth, but this may be because of their incorrect belief that all ideas are equal. Some ideas presented have intentional vagueness, confusion, equivocation or self contradiction and should be criticised and rejected on these grounds.

Non-science disciplines have objective truths. This is obscured by bad philosophy, first in the seeking of final answers and second in the acceptance of incoherent and contradictory ideas (postmodernism).

connections_for_objective_truths_exist_in_all_disciplines_and_finality_in_none_of_them

Knowledge creation is a form of computation

Knowledge creation is a form of computation.

Human minds are universal computers and we create knowledge.

Knowledge creation is a form of computation.

connections_for_knowledge_creation_is_a_form_of_computation

We are at the beginning of scientific progress

TODO

The universe is infinite and any knowledge we have will never be perfect. Thus error correction goes on forever. We will always be finding new things.

TODO.

Dark energy shows how we have more to learn

The universe accelerates as it expands and we haven't solved why.

Dark energy is the name given to the thing causing the universe to expand though we know little about it. It could be used by people as an energy source to power future knowledge creation.

Dark energy is an unsolved problem in cosmology.

We cannot see all the universe

We can only ever reach a part of the universe, but that will be expanding infinitely. It will contain random unlikely things, and our ability to observe those may be limited by how good telescopes can get.

The universe expands at an accelerating rate, so there are parts beyond our reach. However even the parts we can reach expand infinitely. One consequence of this is that the number of unlikely things we can see will increase the more we explore and observe this region of the universe. These unlikely things (anything that can happen will happen) would be observed by large telescopes, and if there is a limit on the size of the telescopes we can make, there may be some upper limit on the sort of unlikely things we can detect (very unlikely vs very very unlikely with larger telescopes).

We see only a measure of the infinite universe. If the cosmological principle (anything that can happen will happen) is right, then the measure of those things that we can observe increases the more we advance. There may be an upper limit on the improbable things we can observe if there is a limit on telescope technology.

connections_for_we_cannot_see_all_the_universe

The problem with anthropic arguments

Some questions seem answerable, but are not because of how they involve infinity. Two are first to ask how likely it is that aliens exist, or second to ask how likely it is that another different sort of universe with different parameters could harbour life. We don't yet know how to mix probabilities and infinities in a useful way.

When we make an observation and note that we are here in this observation, there is a theoretical infinite amount of other situations in which we are not there to make the observation (we exist in a universe with constants compatible with life, or we have life on a planet compatible with life). To ask the question about how likely it is that observable conditions exist in general can cause a problem. The problem arises from the assessment of probabilities when there is an infinity involved.

The Fermi problem (where are the aliens?) and the fine tuning of cosmological constants (why is the universe just right for life?) have problems that affect our attempts at calculations. Any theory involving an anthropic argument must provide a measure for defining probabilities in an infinite set of things. It is unknown how to do this for neither a spatially infinite universe or an infinite set of cosmological constant combinations).

connections_for_the_problem_with_anthropic_arguments

The problem with Quantum suicide

It may be tempting to take actions expecting that in some universe you will end up in a good situation. What you ultimately experience may rely on likelihoods, which are problematic if there are infinite versions of you making that action.

Given that we split in the multiverse, you could make a machine to kill you if your lottery ticket is not the winner. Lottery seems random from your subjective perspective, so you will be a millionaire in all universes where you wake up alive. This is not good reasoning because it requires an additional assumptions:

When making a decision, one should ignore all histories in which the decision maker is absent. Namely, don’t worry about all those dead versions of you. However, similar to anthropic arguments, there are possibly some problems with assessing probabilities of the situation, given there are infinities involved. So it is poorly understood if you can say that it is likely your conscious experience would be to wake up, alive and wealthy.

Making decisions where you expect your conscious experience to continue in some branch of the multiverse is a poorly understood area. It may require more knowledge about how to assess probabilities of infinite sets.

connections_for_the_problem_with_quantum_suicide

The problem with probabilistic simulation theory

Being able to imagine our universe being a simulation running on many computers does not increase the probability that a simulation is likely. This is because we do not know how to assess probabilities of this kind, only probabilities related to the history of our current universe.

Consider the experiment where the universe is like puff pastry and some regions make more instances of you (Oxford 1 instance and London 1 million). You wake up in the dark, should you bet London because there are millions more instances of you waking up in london? No because counting instances of oneself is no guide to probability. So if the simulation of the universe was being run 1 million times, does it make it more likely that we are in one? No for the same reason. We have no knowledge for how to assess probabilities of this kind, we can only assess the probability of different histories, as measures according to quantum theory.

TODO (refine 18 III 34:00, I think it is actually about the uncountable infinity. If the simulation is being created by a future version of ourselves, then there are an uncountable infinity of those futures where a simulation is being run. So probabilities break, even if they each run many simulations, they cannot be counted… Or is it that there are nested infinities?)

Simulation theory probability cannot be assessed by invoking additional instances of the simulation. To do so involves assessing probabilities of potentially infinite instances.

The problem with morality of simulation theory

If someone made our universe as a simulation, they created suffering. It might be that advanced civilisations have advanced morals that would prevent them from doing this.

The simulators of a simulation would have to be okay with creating suffering. It is reasonable that moral knowledge may increase with other advances in knowledge and that because morality is objective, they would choose not to create the simulation.

A society advanced enough to create universes may have moral knowledge that compels them not to create suffering.

connections_for_the_problem_with_morality_of_simulation_theory

The rate of innovation after the singularity will not be too rapid

Technology that helps us learn faster has and will continue to improve our rate of learning. So we are likely able to keep up with new knowledge being created, even if it is from a computer based intelligence.

It has been a persistent belief that the current rate of progress civilisation is making is too rapid and causing too much disruption. Our capacity to enjoy changes in our lifestyle, norms and ethics has been increasing too. We will continue to create technology that allows us to absorb information faster through new interfaces. With our minds being universal computers, we can fundamentally have the same capacity to understand as an AGI.

Our capacity to absorb new information grows and will continue to grow as the result of things we create. An AGI that creates new information will be equivalent to other times in our history when the rate of knowledge creation has increased rapidly.

connections_for_the_rate_of_innovation_after_the_singularity_will_not_be_too_rapid

Humans have previously coped with singularity-type events

In the past we have made progress that has dramatically changed society for the better in a short period of time. Hence it is likely we will be able to do it again if we build an AGI.

The industrial revolution, the agricultural revolution and the enlightenment were all events characterised by dramatic change to society. If we enter into a new phase where knowledge is created at increased rates, then we will likely be able to successfully navigate that too.

That history has had many constructive revolutions suggests that any large advancements we make will be survivable.

Predictions are possible without prophecy

You can make predictions as long as new knowledge is not likely to be created. If something be invented and disrupt what you are predicting, then it is prophecy.

There is a way to tell the difference between prediction and prophecy. It is important to know if new knowledge can be made to affect you are trying to predict. If new knowledge can be made it is prophecy, if new knowledge cannot be made in time then it is prediction.

If you make a concealed hole in your driveway, you can predict the car of a visitor next week will fall into it. However, if a visitor is coming in 5 years, you can only prophecy they will fall in, for new knowledge could be created and your visitor may have bought a different vehicle, one with advanced road scanning, or one that flies. The prophesy-prediction transition point will likely happen sooner than if you were to make predictions about the appearance of the moon, because there are many people working on car technology, and fewer people working on moon-scape or moon-city technology.

The time beyond which scientific prediction has no access is different for different phenomena. For each phenomenon it is the moment at which the creation of new knowledge may begin to make a significant difference to what one is trying to predict.

connections_for_predictions_are_possible_without_prophecy

Speculation is thoughts that lead to conjecture

To speculate is to create a new idea about a current problem that can later be criticised.

Speculation is wondering about what could happen and then creating a new idea about that. Speculation is not prophecy because you are not claiming or expecting that something will happen. It is an idea about a problem and is critical in the path to solve problems.

To speculate is to think about a problem and undertake conjecture formation. This conjecture may then be refuted and in this way knowledge is created.

connections_for_speculation_is_thoughts_that_lead_to_conjecture

People are morally at the center of the cosmos

Some think that people are as significant as rocks, but consciousness makes us more significant. Some think that people are as significant as animals, but we create explanations and that makes us more significant. This is because we can find and fix errors that cause suffering.

It is the concerns of people, rather than all conscious creatures, that are at the center of morality. Some claim that all conscious creatures have a claim at moral status, to the extent that they are conscious. This is incorrect because the sort of consciousness that people have (our capacity to create explanations) is categorically different from all other creatures on Earth. We can make moral advancements.

It is not mere consciousness that makes people special, it is our ability to create explanatory knowledge. This makes us central to morality because we, and not animals, can make moral progress.

connections_for_people_are_morally_at_the_center_of_the_cosmos

There is nothing about this book that needs to be believed

Everything in this book can be modified, challenged or ignored and none of it should be believed. What is here are ideas that can be used and tried out. If there are some truths, then they will tend to survive in your because they are useful.

All that is here is fallible explanations. Any of this can be overturned, criticised and reinterpreted by you in a way that suits you, works for you, and allows you to make progress. To come to a consensus in order to resolve conflicts. A reader should not to adhere to or believe this text in any way. Ideas may imperfectly replicated in the mind of any reader who engages and comes up with the same idea through creative conjecture.

This book contains fallible ideas. If there are some truths here, then they may survive in the mind of the reader who will create the idea again and try it out.

connections_for_there_is_nothing_about_this_book_that_needs_to_be_believed

The problem with atheism based on evidence-based reasoning

Don't start with evidence (I don't see a God) and build ideas (there is no God) around that. Start with ideas (the universe runs on laws) and then see what evidence those ideas require (evidence for the laws). Then notice that no good explanations require God.

People consider themselves atheists because there is no evidence for God. So one should not believe in God. This is wrong because there is no evidence for anything. First seek good explanations and see what the explanation requires.

It is better reasoning to think that there is no God because no explanation requires you to invoke the existence of God. Also recognise that all ideas are fallible, including “there is no God”.

Atheism should come from seeing that no good explanation requires God. It is empiricism to look for evidence for God and then conclude that there is no God.

connections_for_the_problem_with_atheism_based_on_evidence_based_reasoning

There is a lot to preserve in religion

Religions have been a part of our changing and open society. If you got rid of them quickly, there may be unforeseen and terrible impacts on society. Any changes would be better implemented as slow steps.

Religions are institutions and similar to other institutions in society, they are long-running and complex systems that we do not fully understand. There are many positive things that religions may be doing, and care should be taken not to hastily dismantle them. That could cause great harm in unknown ways, including causing society to become more static rather than dynamic.

Religions, like other institutions, may contain valuable truths that benefit society and should be slowly, not rapidly, changed. Their implicit knowledge in their practices may be preventing static society.

connections_for_there_is_a_lot_to_preserve_in_religion

Seek good explanations

TODO

All we can do is choose an infinity of ignorance or of knowledge.

TODO.

Common sense realism is both simple and advanced

Most people are like pilots, flying the aeroplane of their own lives. The study of how things work, how progress is made and how problems are solved are not important to fly the plane. However, the knowledge for how that plane works under the hood is there. Other ways of thinking can tend to reject the idea that there is a real plane to be flown and can get people lost in theory disconnected from reality.

The most rudimentary way to see the world is that things around us are real, and we get on with life knowing that. There are also ways of thinking about the world that diverge from this (sophisticated intellectual idealism, Bayesianism, scientism and rationalism), and they separate people from actual practice of solving problems. Popperian-style epistemology (and subsequently Deutsch/BOI) is another level, where common sense realism is again embraced, but it has the vocabulary to explain how thinking works, how progress is made and how problems are solved.

Popperian epistemology makes explicit what is already implicit in the world. It is a vocabulary for describing the reality that most people already utilise. Other kinds o f epistemology embrace ideas that reject concrete reality.

Good epistemology is a defence of indoctrination

The study of how knowledge is created is important because there are some ways of thinking that do not lead to decisions that are helpful. If you are unaware of how they are wrong, then you may mistakenly think they are good ideas.

The theory of knowledge that Popper and Deutsch communicate is important to understand because it will allow incorrect ideas to be identified. There are smart people in important domains who have frameworks that sound compelling but lead to bad decision making. By understanding how knowledge is created, you can avoid adopting ideas that would otherwise have been compelling.

The study of Popperian epistemology helps inoculates against bad epistemology.

We are at the beginning of infinity and can choose our direction

The universe continues onward, and our future looks good if protect our ability to create good explanations about how things work.

We are at the beginning of infinity and to survive in the long run and make progress we must act. Our actions determine what that infinity looks like, and the outcomes can be:

Ignorance or knowledge
Wrong or right
Death or life

Our ability to create explanatory knowledge is critical in solving problems.

The universe continues on into infinity. If we are to persist and thrive, our actions to cultivate the growth of explanatory knowledge are important.

A dissection of Moloch - Ceramic exploration

Moloch is one counterfactual out of many

Moloch is about something that could have happened.

Moloch (coordination failure) is a term to describe a pattern of problems caused by coordination failure. Coordination failure describes something that did not happen, but could have.

Moloch is a retroactive counterfactual for coordination broadly.

Moloch inverts a solution and presents it as the cause

Coordination might have solved a problem. That does not make the cause of the problem
the absence of coordination.

Moloch is a solution that could have been used. However it is often presented as the cause for something.

An example might be to say the cause for rose garden workers having cut hands is the absence of gloves. Gloves are one solution, but so are handheld tools, or robots. The real cause is that bare hands are being used to manipulate sharp thorns.

A more applicable example is a new invention that causes some problem that affects multiple competing agents. The real cause for the problem is the new invention. The solution may be another invention or creative solution that has nothing to do with coordination.

Moloch is the misattribution of a solution as the cause. It can obstruct the path to finding the best explanation.

connections_for_moloch_inverts_a_solution_and_presents_it_as_the_cause

Moloch is easy to vary

Moloch can be morphed into a variety of shapes.

Moloch (coordination failure) can describe a great variety of counterfactual situations and scenarios. It is easy to vary some coordination scenario that fits a particular setting. If that scenario is criticised, another can be created and so on.

As it is easy to vary, that makes it hard to test. Which makes it a bad explanation.

Moloch is a bad explanation because it is easy to vary.

connections_for_moloch_is_easy_to_vary

Moloch is always invocable for multi-agent problems

If people are involved, coordination can always be imagined.

Whenever there are multiple competing agents involved in a problem, one can invoke Moloch. This is possible even when there is a problem with a very clear and specific cause.

Some children are playing with a ball and one falls on a piece of glass despite someone else having seen it before. Why did this happen? The best explanation is the simplest: “someone left glass on the ground, the children were busy playing and then one child fell on it”.

It is not “the children failed to coordinate and communicated the glass to each other”. The solution is not clearly “better systems for children to coordinate communicate dangers”. The solution could be a single child put the glass in the bin when they saw it, or a groundskeeper starts inspecting the yard before the school day.

Coordination is a general term that can always be applied to multi-agent systems.

Coordination without coercion is difficult

It is not always easy to get people to agree on a topic.

People compete using a variety of ideas and beliefs. Getting people to come together on a topic can be very challenging. It takes a lot of reasoning and effort to coordinate without forcing or coercing other parties.

Peaceful coordination requires significant time and effort.

The misconception of Moloch is that coordination is only one of many possible solutions

A coordination plan is one creative idea for how to solve a problem. Many other ideas exist and may
be better.

Moloch is a meme that encourages artificial prioritisation of coordination-based solutions above other kinds of solutions. This is significant because Moloch can always be applied when people are involved.

Yet there are many other solutions that could resolve the problem, and these may be easier. People need to believe that other solutions can exist and then attempt to create them. However, Moloch is presented as the cause, so rather than looking at the broadest and most comprehensive view of a problem, they may only see the problem narrowly as a problem of coordination. Their creative efforts might then be restricted.

Problems that feature multiple agents that could coordinated are not trivially attributable to or best resolved by coordination.

connections_for_the_misconception_of_moloch_is_that_coordination_is_only_one_of_many_possible_solutions

That Moloch is a meme does not make it useful for understanding the world

The Moloch idea is spreading, but that does not make it a good idea.

Useful memes tend to survive. They might be useful for particular reasons like they get you on podcasts. Moloch may tap into deep fear thought patterns that make people afraid, which makes them remember the idea and trust the bearer of the tale.

Bad ideas can replicate as memes.

connections_for_that_moloch_is_a_meme_does_not_make_it_useful_for_understanding_the_world

Coordination is not required

If people believe coordination is essential when it is not essential, this can lead to bad ideas
about what to do.

Moloch is a misconception that coordination is required to solve multi-agent problems. That is incorrect because a single independent actor may create a new idea that solves the problem.

Using the bad theory of Moloch in ones explanations for the world may lead to other theories that do not match reality. For example, one might create a policy that requires all scientists in a domain to coordinate. This could reduce the variety of ideas they produce and stifle the creation of new explanatory knowledge.

The belief that coordination is required is unfounded and can lead to decisions that impair progress.

connections_for_coordination_is_not_required

Moloch is not real

Moloch refers to some hidden pattern that does not exist.

Moloch (coordination failure) does not appear in our best explanations for how things work. For every problem attributable to Moloch, there is a better explanation.

For example, a group of competing businesses with the problem of bad working conditions can be explained by describing businesses, unit cost, economies of scale, markets, employees and wages. “Failure of coordination” is not in the explanation. One may have a theory about coordination, but it is not required to explain the situation.

Coordination is a specific idea for a specific problem. One might say that a problem is caused by coordination failure and proceed to outline exactly what the coordination entails. However, one may claim that the problem is caused by the absence of some other idea, system or invention and outline details for those. This might be a better explanation.

In the business example one could claim that it is the absence of a monthly business coordination and accountability council that is part of the explanation of the problem. This is the same as explaining the cause as being the absence of some new cheap material or the absence of a basic income regulation. These are ideas for solutions, not explanations.

Moloch is a metaphor for a process that has no basis in reality.

connections_for_moloch_is_not_real

The Moloch framework is a false dichotomy

There are more choices than to join in or not. People can add a third option.

The coordination framework that underpins Moloch is that there are competing agents who have the choice to collaborate or be selfish. There is no such restriction on real people in real situations. There are an unlimited number solutions that they could think of and try.

To present problems as or “do or die” (join the competition and degrade ones values, or fall behind competitively) is simplistic and a failure of the imagination.

Attributing some problem to coordination failure is a failure of the imagination.

connections_for_the_moloch_framework_is_a_false_dichotomy

Moloch is a prophesy that coordination failure will cause problems

The concept of Moloch is an impossible prediction about specific bad things happening.

On Moloch, Scott Alexander writes “… He always and everywhere offers the same deal: throw what you love most into the flames, and I can grant you power.”

The core of Moloch is an imagined destruction-for-power deal. There is never such a deal when people are involved. They can think of new and creative solutions and are not confined to throw or not-throw.

That such a binary deal (coordinate or not) will “always” feature in future problems is a pessimistic prophesy. One cannot predict what problems we will face because they are dependent on the solutions we make.

That coordination failure is “everywhere” is an easy to vary (bad) explanation that does not explain how Moloch is present, other than “it will be there in some form if you look close enough”. Which parts of a multi-agent system will fail to coordinate? In what way? What is the mechanism for this to keep occurring?

The metaphor of Moloch is used to claim that coordination failure is a fundamental and persistent phenomenon. This is an unsubstantiated prophesy about the nature of future problems.

connections_for_moloch_is_a_prophesy_that_coordination_failure_will_cause_problems

The idea of Moloch is not a harmless framework

The belief that coordination failure is fundamental can cause misdirection and pessimism.

Moloch can cause three problems:

People to fixate on coordination when other solutions can solve a problem
Systems that promote coordination, which is unrelated (or oppositional to) to solving problems.
People to feel pessimism because there is no way to escape this “fundamental flaw” that people are claimed to have. If Moloch is real, humans are destined to cause some galactic scale catastrophes filled with misery.

Moloch is a bad theory. Theories that include Moloch are likely to deviate from reality and will not be as useful for solving problems.

connections_for_the_idea_of_moloch_is_not_a_harmless_framework

There is no Moloch behind the prisoner’s dilemma

Sometimes people design systems with undesirable effects. The system can be redesigned.

is a misconception because people are creative and the world is complex. There are possible solutions that do not require coordination.

A prisoner has a problem (prison). They are offered a bargain (and can take or leave it), but that is a false dichotomy. They could try a third creative alternative, such as tell a journalist about the absurd bargain they were offered, which may cause prison reform.

The prison system architect is faced with a separate problem, the emergent behaviour of the prisoners is creating outcomes that do not match the moral code of society. The goal of finding information from prisoners and ensuring future reduction in crime is not being met by offering out “faustian bargains”. The prison system design is the real explanation for the problem, not coordination.

Prisoners dilemmas are an artificial constraint that makes it seem like some situations are doomed to impossible coordination tasks. In reality, systems are not isolated and people can “break the rules of the game”.

A system may be designed with errors that result in problematic dynamics (prisoner dilemmas). A better system can be designed.

connections_for_there_is_no_moloch_behind_the_prisoners_dilemma

There is no Moloch behind dollar auctions

The problem with dollar auctions is not how to bid, but how to avoid getting stuck in the auction.

Two people end up in an auction that favours the auctioneer. It is argued that this points to a real coordination problem humans will tend to face.

The misconception is that coordination failure is not the best explanation for the dilemma. The real explanation involves looking at the broader system. How did they come to be playing this rigid game? Why are other systems of auctioning not being used? Is there monopolistic capture of auctions? Those will give the real explanation of the problem.

Dollar auctions are not Molochian, the scope of the problem includes the design and instantiation of the auction itself.

connections_for_there_is_no_moloch_behind_dollar_auctions

There is no Moloch behind the fish farming story

When groups want something enough, this signals a demand for a new and creative opportunity.

A community of fish farmers all want everyone to use filters, but don’t want to themselves. This is their vocation, and so it is very important to them. There is no Moloch here, because people are able to come up with creative solutions to their problem of how to make the most money.

The solution space includes anything physically possible.

One farmer may downscale their fishing and offer a “filter inspection service”. They frequently go around filming the farms. Other farmers can pay a small fee to see the footage. Knowing they are being observed, the farmers may all converge on using filters.

Problems that are large and significant are opportunities for creative solutions. These solutions are changes to the system, which are always possible if the laws of physics allow.

connections_for_there_is_no_moloch_behind_the_fish_farming_story

There is no Moloch behind the Malthusian trap

It does not make sense to point to animal systems as evidence that humans will be caught in
traps. Humans can invent ways to get out of traps.

A population of rats go to a paradise island, which results in boom-bust cycles of plenty and scarcity of nuts. The rats are said to be doomed to these extremes as a consequence of Moloch.

This analogy does not apply to people, who can think about the problem overall and come up with solutions for different systems for getting and distributing nutrition.

Moloch is also not a guiding principle for evolution. Islands are complex and emergent phenomena, and one might see a growth in a population of rat-eating birds. This might indeed result in an equilibrium where the rats have a stable population with enough food.

The Malthusian trap incorrectly imposes the rigid constraint that people cannot create new explanatory knowledge.

connections_for_there_is_no_moloch_behind_the_malthusian_trap

There is no Moloch behind capitalism

Capitalism generates solutions to real problems. Solutions cause new problems.
The trick is to apply capitalism to those new problems, not abandon capitalism.

It is argued that a competitive sweatshop garment industry is forced to abandon all values except profit. This means that many people have unpleasant lives in this industry.

This is indeed is a problem, but it is not an inevitable “driving force” of capitalism.

First there was a problem of expensive clothes. Someone solves this problem by relocating to a country with lower wages. A real problem was solved. Now there is a new problem, that people have low incomes.

Now because the industry has a large scale, it will attract the attention of many people with new ideas. A premium brand label that has ethical workplaces. An invention to improve air quality that can be sold to many factories for a low unit price. An invention that a worker can use to work faster, easier and protect their hands.

One should not conflate solutions to problems as being problems themselves. Garment factories solve scarce clothing, and new and interesting problems arise. Capitalism causes problems, but only because it solves problems first. Seeing capitalism as a problem generator rather than a solution generator is a misunderstanding.

Capitalism is the consequence of people working hard to solve real problems they see today. Their solution is unpredictable. New problems will emerge, and these too will be unpredictable but solvable.

connections_for_there_is_no_moloch_behind_capitalism

There is no Moloch behind the two income trap

Women were not allowed careers. When that problem was solved, many lives flourished.
A new problem of expensive houses arose, which is a new and addressable problem.

It is argued that if a suburb expensive houses has one-income households, eventually everyone will have two-income households, but the houses will be more expensive too. Moloch is said to be the cause of this undesirable end state.

People get jobs for many reasons including community, life meaning, expression and agency in the world. 50% of the adult population created visions for their career and actioned them. This is not trivially reducible to “they did it to get a house”.

A happy and fulfilled two-income household can afford to pay more for a house they want. If many such people exist, the suburb culture one of happy working people, doing productive things for each other, living in costly homes. Costly homes but happy people. A career is a solution a person has to life meaning, expression and agency in the world.

A person who does not like their work does have a real problem but can search for solutions that match their values such as a new career or a move to a new suburb.

The cause of a problem (more people working) may be something that is the solution to a different but very significant problem (cultural barriers to the opportunity to work).

There is no Moloch behind the agricultural revolution

The invention of farming caused many problems. People prefer the new problems to the old problems.

It is argued that hunter-gathers in a sufficiently intense competitive environment are driven by Moloch into an undesirable end point of agriculture. Diseases, oppression and pestilence, where before they had none.

However, hunter-gatherer people had real problems. Severely unpleasant, terrifying and life threatening problems that today we do not have. Agriculture is an invention that directly solved food scarcity and gave people time to work on other problems, leading to writing, creating, inventing and sharing ideas.

Hunter gatherer societies did not have access to the best ideas of how to stop mysterious diseases, or dethrone an oppressive village elder or stop strange groups people speaking unfamiliar language from attacking.

The problems agriculture caused were new, emergent and unpredictable. The problems agriculture solved were real problems that real people wanted to solve.

The solution to a problem causes new problems. The solution is not a failure. New unpredictable problems will always arise and are not grounds for arguing that problems should not be solved at all.

connections_for_there_is_no_moloch_behind_the_agricultural_revolution

There is no Moloch behind arms races

Two countries that never fight but dedicate money to weapons accidentally invent things that are
useful throughout society.

It is argued that spending money on military is bad, for different countries could coordinate to all spend lower amounts if Moloch was driving them to be the biggest spender.

First, military spending results in real people applying themselves to create new things that are tested against reality. This focus of reality testing (it must actually work), means new explanatory knowledge is created frequently. This has reach beyond military domains. So one can see that military spending directly leads to innovation in other fields we value (medicine, computing, etc.). So one may end up with unused missiles and spinal implants for paraplegia.

Military solves real problems. One of the most harmful things is a static society that prevents criticism and change. It is not easy to defend against a culture who is not open to correcting errors and who wishes to spread that culture. Those cultures have existed and one real solution is a military. Once all societies are dynamic and can internally correct and improve, militaries may be required less.

An arms race is the dedication of vast resources to people who create new and unpredictable explanatory knowledge that has reach in non-military applications. This is not wasted resources.

The creation new knowledge is not containable to a single domain. All serious research efforts produce knowledge that can solve unrelated problems.

connections_for_there_is_no_moloch_behind_arms_races

There is no Moloch behind cancer

A cancer cell in a body is part of a broader system that uses blind guesswork to discover reality.
That sometimes this results in bad ideas doesn't mean Moloch is driving things to fail.

It is claimed that a cell may end up causing a situation that is bad for itself due to the fundamental presence of a Molochian process.

A cell is an expression of a genome, which is a system for blindly experimentally generating and evaluating random ideas and seeing if they match reality.

In the example a cancer cell is perhaps part of an experiment that explores what is a good amount of mutation to permit. Ideas (new gene mutations) for the next generation can be tried faster with more mutability, but at the expense of stability once the body is grown (cancer). It is genes that are the replicating unit in evolution-based knowledge creation, not cells. So seeking “rational” behaviour at the cellular level results in a meaningless analogy and coordination between individual cells is not required to solve the problem.

Looking at the broader system, the presence of organisms with tumours that arise after reproductive age could indicate some knowledge about reality. For example, the rate of high energy solar radiation particles on Earth, and the presence of an ever-changing environment of shifting rewards and dangers. That tumours sometimes kill might be the optimal setting for biological systems optimised for blind adaptation.

Systems that create non-explanatory knowledge may have units that don’t make sense if viewed in isolation. Overall, such systems are not doomed to fail due to some fundamental Molochian force.

connections_for_there_is_no_moloch_behind_cancer

There is no Moloch behind a race to the bottom

Competition drives the creation of new ideas and helps us understand what it is that we really
value.

It is claimed that competing agents will tend to sacrifice a common value and become less scrupulous over time. This is proposed as a general truth, which can be considered Molochian as unless agents coordinate, common values will disappear.

One specific example is that a government will lower taxes to levels to attract people, but that this overall causes tax revenue to be too low and no change in the number of businesses. The mistake here is imagining that competing governments must only choose between making taxes high or low. They can of course try completely different ideas instead, in conjunction with, or after the tax high or tax low decision.

In general error of this framework is thinking that the competing agents only have two potential actions, to copy behaviour or lose. Agents can create innumerable new and novel creative ideas that can be tried instead, alongside or after a given set of options. There is no single “bottom” of some Molochian pit that all agents are destined for.

One example of a creative idea a government can try is to keep high taxes, but launch a campaign showing off its wonderful public infrastructure and health care, and make its spending transparent and open to criticism. Employees might find this appealing. Businesses might locate where they can find the best employees. Coordination is not required.

There is no race to the bottom trap for multi-agent systems involving people. This is a false dichotomy, for agents may choose one or many novel paths of action. There is no guaranteed doom as long as agents are free to try new things.

connections_for_there_is_no_moloch_behind_a_race_to_the_bottom

There is no Moloch behind education incentives

Some people can see an expensive degree is broken a metric. They can use this to their advantage
to outcompete others who do not see this reality.

It is argued that there is Molochian pressure that drives people to pay for high status degrees that do not deliver useful skills.

The mistake here is in thinking that as a metric (a degree) starts to lose value, that people will not see this and use it as an opportunity to solve their own problems.

An employer might correctly see that graduates have very little useful skills and have a lot of debt. These employers will have a competitive advantage because they can find highly skilled low-status degrees or self-taught people that are happy to with a lower income.

A high status educational institution might detect that it’s degrees are starting to signal “low skill”, and could create ideas to resolve that. Coordination is not required.

Metrics that fall to Goodhart’s law become bad theories. People are not forced by Moloch to stick with old metrics, and can come up with better ideas for metrics that will have a competitive advantage.

connections_for_there_is_no_moloch_behind_education_incentives

There is no Moloch behind scientific research

Science produces real value for the world. This is a large incentive to improve the way science
is conducted. Those who solve this problem have a lot to gain.

It is claimed that there is a Molochian force that drives science to produce confusing, late, biased and misleading content. People can see what must be done but cannot get it done.

The mistake here is the claim that people cannot create new ways to solve these problems. There is extraordinary wealth that comes from the creation of new explanatory knowledge. It is reasonable that people will create new ways to increase the quality and rate of science. Those who do could have access to part of that value.

For example, retroactive funding for discoveries that resulted in material technologies could make some basic science researchers extraordinarily wealthy. This could create a cultural shift focused on fast, broad, real and novel science. A small isolated group could implement this, and coordination is not required for the idea to spread.

The current problems in science arose after much far more significant problems were solved. Science used to be isolated, unpublished and unreviewed. People knew these were problems. Individual people solved these problems through things like, the invention of the book, the printing press, the journal and the internet.

People that see better ways of conducting science will produce more good explanations than others and have a competitive advantage. This knowledge of how to do science will tend to persist.

connections_for_there_is_no_moloch_behind_scientific_research

There is no Moloch behind government corruption

If a government is corrupt, the democratic process allows them to be removed, as long
as people agree that it is a real and worthwhile problem to vote on.

It is claimed that government donations to corporations is inevitable because corporations can affect government officials.

For this argument let’s assume that giving money to corporations is bad.

If the amount of money is truly large and could be repurposed to things voters really care about, this an opportunity for a party make this information known to the voters and to put forward ideas about what should actually take place. If voters think it is a good idea, then this could plausibly get the party elected.

If the majority of a population disagree with the experts about the corruption, then the specialists can solve that problem by presenting their rationale to the wider public more creatively.

A democracy is a system that the capacity to remove bad rulers. A party with corrupt spending practices can be replaced by one with different ideas if enough people see and agree.

There is no Moloch behind the US Congress

A good government will find and debate things people disagree about. That may not be nice to
watch, but it is how progress is made.

It is argued that the low approval rating of the US Congress is a manifestation of Moloch. People should be happy with the way Congress it is handling its job.

The mistake here is in thinking that disagreement about the action of Congress is a good explanation for its failure.

In a democracy, parties debate current ideas that people are passionate about. There are an enormous amount of old ideas that both parties agree on and which do not get debated any more. Only new ideas that the people care about and also disagree about are subject to debate. When operating well, parties will find the core problems that society is wrestling with. These ideas will be the most controversial, where there is no clear majority opinion. That is, the idea divides the country in two.

Congress represents these divisive ideas. One day society may come to partly agree about aspects of these ideas and will move on to new and divisive ideas. A population cannot ever fully agree.

That is to say that representative democracy, when working well, will represent divisive ideas. The theory that everyday people should feel approving of congress overall is a misconception. A better theory might be that if everyone is happy with congress, then they are not finding the real issues that people find important but disagree about.

A democracy that functions well is expected to be representing the most divisive ideas. Disapproval of the US Congress implies it is working on the most important areas for people.

connections_for_there_is_no_moloch_behind_the_us_congress