David Deutsch’s worldview

Reality is comprehensible - four fundamental disciplines, all necessary and interdependent to understand the world

  • Quantum physics is the deepest language of physics in which all other theories are laid out and the basis of understanding our physical universe
  • Epistemology is the theory of what constitutes knowledge and how to acquire it
    • A good philosophy of knowledge is the cornerstone of human knowledge creation
    • Knowledge is information that is useful, it has the property that it tends to keep itself instantiated
  • Evolution is the theory of what processes in nature keep themselves into existence
    • It is the theory of replicators (genes and ideas in particular)
    • It explains how large objects - life and thought processes creating knowledge - can be understood in terms that do not follow from their low-level definitions in terms of physics (emergence)
  • Computation is the theory of what processes in nature are independent of the material substance that they are embodied in
    • Turing’s universality of computation says that, conceptually, there is only one way to physically process information
    • If we consider the application of the laws of physics as computation, this implies that a physical computer can (in principle, given enough resources) simulate physical reality, and so can the collective of human brains along with our tools (including computers)
    • This is the best explanation for the fact that physical reality can be understood by itself

How are the theories interdependent? The central narrative is that our physical universe is on the path to understanding itself through human knowledge creation

  • The creation of knowledge - first by biological evolution until humans (life), then by idea evolution (thought) - explains how this universe comes to understand itself first by genes getting organisms to pass them on by reproduction and then ideas getting humans to pass them on by dialogue
  • Computation explains why this is possible through the Turing principle, justifying optimism about the beginning of infinity: humans are on the path to creating infinite knowledge and impacting the universe (multiverse) more than anything has ever before
  • Note that for this to happen, the fabric of reality should be and is layered for easy self-access: life and thought evolution processes must be efficient, they cannot require too much time or other resources

In other words, human reach is the same as the reach of explanatory knowledge and is thus bounded only by the laws of nature

  • If there were something that couldn’t be explained, this would have an explanation (be a law of nature or a consequence of one)
  • Saying otherwise is saying that the world is inexplicable - we might as well invoke the Gods of ancient myths
  • There are two claims
    • Hardware: “human brains are universal” - Turing principle, fairly non-controversial, albeit not widely internalized
    • Software: “human explanations are universal” - Deutsch makes a strong case for this

This worldview is fundamentally optimistic - it places human minds at the center of the physical universe, and explanation and understanding at the center of human purpose

Only infinite progress by knowledge creation is sustainable for our species to thrive

  • Problems are inevitable, and the greatest dangers are unforeseen (e.g., global warming in the 70s)
  • But soluble - every transformation that is not forbidden by the laws of physics is achievable given the right knowledge
  • If we fail at anything, it will be because of knowledge that we fail to create

Extension of Popper’s definition of knowledge creation

Theory (explanations) comes before observations - acquiring knowledge is a creative process to come up with our best guesses to explain reality and iteratively criticize them

  • Good explanations address all current criticisms and explain phenomena better than alternatives
  • They make testable predictions - more broadly, beyond Science, in philosophy or mathematics, they are criticizable
  • But it’s not only about predictions but also about the explanation itself - it must be hard to vary: all of its components are necessary - note that good genetic adaptations, like good explanations, are hard to vary while fulfilling their functions
  • Explanations that are hard to vary are hard to come by, and often make precise and risky predictions

This definition of knowledge creation explains the need for emergence: independent levels of understanding like atomic physics, fluid dynamics, or human behavior

  • Particle collisions explain everything, so they explain nothing, just like God explains everything, so it explains nothing, they’re not useful levels to operate at
  • We shouldn’t discount fundamental theories that emerge as good explanations just because we can’t lay them down in the language of physics; emergent explanations are very often the only known good explanations
  • The idea of cause itself is emergent: there is no concept of cause in the laws of motion of elementary particles (the position of a particle in the past is determined by its position in the future just as much as the reverse, this is not what we usually mean by cause and effect); we cannot perceive causation, just a sequence of events (David Hume)

The most important limitation of knowledge creation is that we cannot prophesy: we cannot predict the content of ideas yet to be created, nor their effect

A bad philosophy is a philosophy whose effect is to close off the growth of knowledge:

  • Empiricism (and pushed to the extreme, positivism) is the theory that we gain knowledge only through observing the world - this makes no sense: without formulating a conjecture, we don’t even know what to observe or how to interpret it
  • Instrumentalism is the theory that Science is only about predictions - this makes no sense: even if we had an oracle that could give us the outcomes of any experiments, we wouldn’t know how to use it without formulating hypotheses/explanations to decide what experiments to run

Ideas (like genes) evolve via iterative competition by variation and selection

Knowledge is information that is useful, information that can affect the physical world towards an objective. It is a replicator: it has the property of keeping itself physically instantiated. There are only two types of knowledge:

  Evolutionary Explanatory
Unit of selection Gene Idea/explanation
Storage place DNA of living organisms Human brains and behaviors
Variation Random mutations New creative explanations created intentionally for a purpose
Imperfect communication from one mind to another
Communication Sexual reproduction of holder’s organism Communication from holder’s mind to another mind
Selection Competition with other genes
Via survival and reproduction pressure from the environment		 Competition with other ideas/explanations
Rational ideas via criticism and experimentation in a dynamic society
Anti-rational ideas via conformity and dogmatism (disabling critical thought) in a static society
Reach Low: usually local to an organism and environment High: our best explanations explain well beyond the phenomena that prompted them, they can even have unlimited reach
Speed of evolution Extremely slow
Blind random variations
At the pace of generations
Credit assignment problem		 Extremely fast and accelerating
Intentional variations
Many times in one generation
Can benefit from human tools for faster variation, communication, and selection: personal computer, Internet, personal AI
Implication for organisms Factory for converting resources of a fixed type into more such organisms Human bodies (including their brains) are factories for converting anything into anything that the laws of nature allow (universal explainers and thus universal constructors)


For most of human history before the Enlightenment, societies were largely static:

  • They had customs and laws suppressing criticism of the status quo and creative and critical faculties of their members by valuing conformity, obedience, piety, and devotion to duty
  • They evolved at the pace of centuries, and even worse, the selection of ideas was flawed: false and harmful ideas were likely to be propagated because conformity was valued over criticism
  • The Enlightenment was characterized by a rejection of authority, dogma, and any system of belief that insulated itself against criticism and change - this led to an explosion in explanatory knowledge

Since then, modern dynamic societies have embraced change, creativity, and criticism:

  • Ideas with the best chance of survival at scale across varied contexts are truths with reach (like the laws of physics), which survive criticism
  • Given knowledge acquisition is an error-correcting mechanism, not destroying the means of error correction via criticism and experimentation might be our best basis for morality (optimizing for humanity’s flourishing) - doing so would be negating our biggest advantage as a species

How did creativity - defined as the ability to come up with explanations - evolve while societies remained static? Deutsch’s explanation:

  • It is necessary to propagate existing ideas and behaviors, not just to create new ones - humans explain other’s ideas and behaviors to learn them; they don’t just mimic like other species (apes, parrots); thinking otherwise would be the same misconception as empiricism
  • In static societies, it was used to infer intentions behind behaviors to more accurately enact the prevailing dogmas
  • This ability of human brains to explain evolved to propagate ideas, becoming more and more universal, and ended up enabling the creation of new ones - it is necessary and sufficient to propagate ideas and has an incredible reach because it enables the creation of new ones
  • This is what makes humanity special: we are creative universal explainers - it is partly hardware (genetic adaptations of brains) and partly software (an appreciation for creativity and criticism)
  • Given explanatory knowledge evolution is incredibly more efficient than evolutionary knowledge evolution, recorded history is largely the history of ideas, and so will the future of the universe

Humans play a special role in the universe

  • To understand humans and everything that happens on Earth (why Einstein is taken to Sweden and given some gold), aliens would need to understand everything there is to know about the world (general relativity, philosophy, morality)
  • Knowledge impacts the physical world more than anything else, thus the things that create knowledge (evolution and humans) are necessary in any explanations of the world; this is the definition of being significant

The nature of mathematics

The prevailing view in mathematics is that when we have proved something, we know with absolute certainty that it is true. But this is not the case: we cannot guarantee that a proof that was thought to be valid will not one day turn out to contain a profound misconception, made to seem natural by a previously unquestioned assumption about the physical world, the world of mathematical abstractions, or about the way they are related

At the turn of the 20th century, David Hilbert published a list of 10 problems he hoped mathematicians would be able to solve over the century. The 10th problem was to find a set of rules of logical inference with the property that they were provably consistent - we would be able to prove that only true statements could be proved with these rules using the rules themselves

In 1931, Gödel proved this 10th problem is insoluble

  • He first proved that any set of rules of inference capable of validating even the proofs of ordinary arithmetic could never prove its own consistency therefore there is no hope of finding the self-consistent set of rules Hilbert envisaged
  • Second, he proved that any sufficiently rich - able to derive ordinary arithmetic - set of rules of inference is incomplete: some statements are true but unprovable through these rules

Proof is a physical process

  • It is a type of computation, with ink on paper, a human brain, or a computer, that models the properties of some abstract entity and whose outcome establishes that the abstract entity has a given property
  • This physical nature of proof is at the center of Gödel’s work, which establishes a correspondence between proving mathematical statements and numerical computation - see Douglas Hofstadter’s work

The confidence of mathematicians in the absolute certainty of mathematics stems from a misconception between the subject matter and the methods of mathematics

  • Subject matter: properties and relationships of abstract entities, which are absolute truths
  • Methods: physical computation, subject to misconceptions

Explanation plays the same role in mathematics as it does in Science

  • Explaining and understanding the world - the physical world and the world of mathematical abstractions - is, in both cases, the purpose of the exercise
  • Proof and observation are just the means to check our explanations
  • The usual order of events in mathematics research is for the mathematician first to understand something about the abstraction in question, then to use that understanding to conjecture how true propositions about the abstraction might be proved, and finally to prove them, not the other way around

The multiverse

Quantum phenomena (the results of the double-slit experiment) cannot be explained by the events that we see in our universe alone; they can only be explained by a “ghost photon” in a parallel universe passing through one slit while the photon in our universe passes through the other

  • In fact, a photon is present across an infinity of parallel universes: there is a field (or waves) in the multiverse for every particle that we observe in one universe, the particle has multiple positions and multiple speeds - this is called the uncertainty principle
  • In this experiment, the photon’s paths taken in two universes split to go through the slits to later recombine - this rejoining of histories is called an interference event, it happens either very soon after the split or never
  • The fact that results change when we put a detector at one slit can be explained by the fact that this causes the two universes to diverge (in one the detector fires while in the other it doesn’t), which prevents interference - this phenomenon is called decoherence
  • The reality that we see and have traditionally thought of as the universe is a tiny sliver of the whole of reality (an infinity of parallel universes: the multiverse)

David’s view is still a minority in physics because others employ bad philosophy (epistemology) and retreat to instrumentalism (let’s not try to explain and just make predictions)

  • This might seem counter-intuitive at first but it is our best explanation
  • The whole project of Science is about explaining the seen in terms of the unseen (dinosaurs, the Big Bang)
  • As we better understand the world, its size keeps increasing - from regions of the world (the New World), to our full planet, to our solar system, to our galaxy (the Milky Way), to billions of galaxies, to an infinity of parallel universes?

Random insights

About universality:

  • The jump to universality: the tendency of gradually improving systems to undergo a sudden large increase in functionality, becoming universal in some domain
  • Initially, the genetic code and the mechanism that interpreted it were both evolving along with the organisms they specified, but there came a moment when the code stopped evolving (DNA) yet the organisms continued doing so
  • The jump to universality of DNA happened when it was still coding only for bacteria, yet it was from this moment able to specify all species, including human brains - what exactly DNA is universal for and why is still unknown
  • The most significant universality is that of humans as universal explainers, which makes us universal constructors as well

About utilitarianism:

  • Moral philosophy is about what to do next: what sort of life to live, what sort of world to want - how to treat others is a subset
  • Utilitarianism does not make sense because it is circular: a human’s preferences are defined as a function of their ideas of what sort of life to live and what sort of world to want, thus we can’t define these in turn as a function of human preferences alone
  • Utilitarianism played the same role as Empiricism did in epistemology: it acted as a liberating focus against prevailing dogma, although it contains little truth in its own positive content