NNaval
← All episodes
11 August 2023

David Deutsch: Knowledge Creation and The Human Race, Part 2

4Frameworks
12Insights

Frameworks in this episode

Insights & moments

The myth-busts, hot takes, explainers, and tools worth keeping.

Myth Buster· 2

Myth Buster00:00

There is no bucket of final truths waiting at the end of science

Naval opens by naming a misconception common in academic and intellectual circles: that discovery terminates in a finished set of true theories, after which humans relax while machines do the work. He argues this picture of knowledge is not slightly wrong but infinitely wrong, because there is no point at which the explanations stop improving. Deutsch's whole account of knowledge creation depends on the loop never closing.

  • The final-theory picture treats knowledge as a bucket to be filled and then carried around
  • Naval calls the vision infinitely wrong rather than merely inaccurate
  • The intellectually retired future of sun chairs and cocktails never arrives
  • Popper's starting point, that everything begins with problems, rules out termination

people think that there's a final theory that what we're trying to achieve is a bucket full of theories that will be the truth at…

Naval Ravikant · 00:00

it's not just that it's a little bit wrong it's infinitely wrong

Naval Ravikant · 00:30
#epistemology#progress#popper#misconceptions
Myth Buster16:00

The Kuhnian paradigm shift is pure fiction

Naval notes that Kuhn's story of grand revolutions has more support in intellectual circles than Popper's account. Deutsch concedes Kuhn said some correct things but rejects the picture of young iconoclasts waiting for stubborn elders to die. He knows of no actual case that fits it, and points out that the people who create a new paradigm usually understood the old one and were solving its problems.

  • New paradigms are built by people who mastered the previous one
  • Deutsch cannot name a real case matching the generational-replacement story
  • Stubbornness exists, but there is no algorithm making the stubborn person right
  • Lister and Semmelweis resisted changes to working practice and perceived dignity, not just theory
  • The generational story explains nothing, treating theory change like a change in hemlines

it's not the case that we completely do away with the previous Paradigm and the people who create the new paradigm tend to have understood…

David Deutsch · 16:30

the thing is it's pure fiction I don't know of any actual situation where that happened

David Deutsch · 17:00
#philosophy-of-science#kuhn#popper#scientific-revolutions

Hot Take· 3

Hot Take11:00

Wave-particle duality is a way of shutting up the questioner

Deutsch describes physics professors who answer hard quantum questions with the formula that it is a particle and a wave at the same time, then tell students they will get used to it or that the question is not meaningful. He argues the student's question comes from taking quantum theory seriously, not from misunderstanding it. Naval adds the born-as-a-particle, lives-as-a-wave version he was taught, which never explained how the transition happens.

  • Repeating a formula is treating a theory as a form of words rather than a claim about reality
  • Declaring a question meaningless silences the questioner instead of explaining
  • The particle-then-wave-then-particle story has no account of the transition
  • Interference experiments only make sense if the quantum entity is neither particle nor wave

they have learned to say ah well it's a particle in a wave at the same time

David Deutsch · 11:00

is not an answer to that question it's a way of shutting up the questioner

David Deutsch · 11:30
#quantum-physics#teaching#explanation#many-worlds
Hot Take19:30

Take feedback from reality, not from other people

Naval argues that the more feedback you take from other people the more likely you are to go astray, while feedback from reality and nature moves you closer to the truth. He notes that parts of science get tangled up in philosophy and academia where nobody interacts with the physical world, so a social feedback loop takes over. Deutsch counters that knowledge can still grow in philosophy, morality and epistemology, because taking a theory seriously supplies criticism even without a physical test.

  • Social feedback loops drift; reality does not negotiate
  • The rockets do not have to fly in fields where nothing gets built
  • Deutsch resists the implication that non-empirical fields cannot progress
  • Taking a theory seriously is the substitute for a physical test

the more feedback that you take from other people the more likely you're to go astray

Naval Ravikant · 19:30

there is this social feedback loop where you're talking to other people you're not always building things the Rockets don't have to fly so to…

Naval Ravikant · 20:00
#feedback#academia#truth#epistemology
Hot Take29:30

The war on misinformation is the opposite of Popper

Deutsch says the political system exists to solve problems, which requires letting rival theories confront each other and never assuming anyone has the final answer. He finds the current rage against misinformation troubling, especially when Popper is invoked to justify silencing people. Naval adds that every eventual truth is first labelled misinformation by the other side, so eliminating misinformation in advance is impossible for the same reason knowledge cannot be had a priori.

  • Petition and redress, not suppression, is the design intent of the system
  • Popper is misquoted to justify shutting down opponents
  • Error correction needs a way to remove bad rulers and reverse bad decisions
  • Successful new truths look like misinformation because they contradict current belief
  • Knowledge has to be conjectured and discovered, so it cannot be pre-screened

nowhere is it a assumed that someone has the final answer this is why the current rage against misinformation is so troubling

David Deutsch · 29:30

every eventual successful truth is defined as misinformation by the other side because it contradicts what is already believed to be true

Naval Ravikant · 30:00
#free-speech#misinformation#popper#politics

Explainer· 5

Explainer07:30

Simplicity is not prior to science, it is posterior

Naval raises the criteria people use to spot good explanations, including risky narrow predictions such as Einstein's starlight bending, and the preference for the simplest answer. Deutsch rejects the simplicity criterion as usually stated: you can only define simplicity once you already have a theory of physics, because what counts as a short program depends on how bits behave. Quantum computation changed exactly that, making previously complex things simple.

  • Good explanations make risky, narrow predictions nobody would have anticipated
  • Simplicity can only be measured relative to an existing theory of physics
  • Quantum computation redefined which computations count as simple
  • This undercuts Solomonoff induction, which assumes a fixed measure of program length
  • Deutsch expects a future physics to shift the notion of complexity again

so there is no scale of complexity or Simplicity that is prior to physics

David Deutsch · 08:30

Simplicity is not prior to science it's posterior

David Deutsch · 09:00
#explanations#simplicity#quantum-computation#induction
Explainer12:30

Why Einstein did not throw Newton away

Naval asks how we can be getting closer to reality when Einstein removed the gravitational force that was central to Newton's explanation. Deutsch answers in two parts: the theories share a great deal, including the heliocentric picture and the claim that the Sun causes planetary motion, and more importantly Newton's theory contains the problems that Einstein's theory solves. Instantaneous gravity and the dark night sky were live problems inside Newton's system.

  • Both theories say the Sun's presence causes the motion of the planets
  • Curvature of space-time replaces force but keeps the causal role of the Sun's mass
  • Newton's instantaneous action at a distance was recognised as a problem before Einstein
  • Olbers' paradox, the black night sky, could only be patched ad hoc within Newton
  • Newton in turn solved the problem in Kepler's theory that Galileo could not accept

the influence is not a force it's a curvature of space-time but that curvature of space-time is caused by the mass of the Sun

David Deutsch · 13:30

Newton said that gravity travels instantaneously that was a problem which people recognized before Einstein

David Deutsch · 14:00
#physics#scientific-progress#newton#einstein
Explainer23:30

How to work on foundations while rejecting foundationalism

Naval asks how Deutsch can go as deep as possible into foundations while opposing foundationalism. Deutsch answers with the relationship between physics and structural engineering: foundations explain why the higher level theories are as they are, but you cannot build a bridge out of Newton's theory. Christopher Wren succeeded partly by using Newtonian mechanics to decide pillar spacing rather than relying on a master builder's eye, yet nowhere in the Principia is there a suspension bridge.

  • Foundational theories explain higher level theories, they do not replace them
  • Wren used Newton to compute spacing instead of trusting a builder's eye
  • Foundations also serve as a critic, ranking designs as better or worse
  • The suspension bridge was invented later and is nowhere in Newton
  • Newton's real contribution was unifying understanding across fields

foundations are theories that explain why the higher level theories are as they are but you can't use Newton's Theory to build a bridge

David Deutsch · 24:00

nowhere in Newton's principia is there a picture of a suspension bridge

David Deutsch · 25:00
#foundations#engineering#explanation#newton
Explainer25:30

Why the English Enlightenment produced change without revolution

Naval asks why so much came out of England, from Newton to Wren to Turing. Deutsch answers that the Enlightenment there was a rebellion against authority that was not utopian. Instead of abolishing the ruling structure and installing a final truth, reformers extended existing privileges to more and more classes of people, so an Englishman's home became his castle rather than only an aristocrat's. France and Germany reformed by abolishing, and inherited a tradition of utopian institution building.

  • England's rebellion against authority framed problems as fixable, not as enemies to remove
  • Parliament and property rights were widened rather than rebuilt
  • Utopian reform writes rights down once and makes them hard to change
  • Britain produced rapid change over centuries without sudden revolutions
  • In the 1930s British fascism never won a single MP and faded on its own

it was a rebellion against Authority but it was a non-utopian rebellion

David Deutsch · 26:00

it was reforms were all about abolishing things abolishing the Tyrant

David Deutsch · 28:00
#enlightenment#britain#political-philosophy#institutions
Explainer30:30

The crystal of knowledge across the multiverse

Naval summarises the idea from The Fabric of Reality that ties together epistemology, computation, quantum physics and evolution: knowledge is a thing that causes itself to be replicated in its environment. Useful genes, ideas and inventions get copied, while noise does not, so knowledge should appear as a structure common across the multiverse while randomness differs. Deutsch pushes back on the practical payoff, because there is no limit to the size of error we can make, so a long-lived religion can look just like a theory heading towards truth.

  • Knowledge is what causes itself to be copied: fire, adaptive genes, working inventions
  • Junk DNA and noise differ across branches while knowledge is common to them
  • A multiverse view would in principle show truth as what is shared
  • Deutsch objects that error has no upper bound, so size is no guide to truth
  • The single-universe version, judging by persuasiveness, fails for the same reason

knowledge is a thing that causes itself to be replicated in the environment

Naval Ravikant · 30:30

there is no limit to the size of error we can make

David Deutsch · 32:00
#multiverse#knowledge#evolution#fabric-of-reality

Takeaway· 2

Takeaway19:00

Vacuum tubes were obsoleted, not refuted

Naval makes the same point through technology. Computing went from analog attempts to vacuum tubes to transistors, and the analog-to-tube jump was larger than the tube-to-transistor jump. Nobody uses vacuum tube computing now, but that does not make it wrong. It was a necessary stepping stone, closer to the truth than what came before, and there was a great deal to learn from it.

  • Superseded technology is a stepping stone, not an error
  • The size of the jump between stages varies but the direction is progress
  • Technology makes the progress visible because artefacts either work or do not
  • The same reading applies to superseded scientific theories

now we don't use vacuum tube Computing anymore it's been obsoleted but doesn't mean it was wrong it was a necessary stepping stone

Naval Ravikant · 19:30
#technology#progress#computing#innovation
Takeaway33:00

No shortcuts: one false idea can drown the truth for a long time

Deutsch closes by ruling out any bound on error. If errors had a maximum size you would know that after a certain mistake your next idea must be true, and nothing like that holds. Knowledge creates non-linearities, so a single false idea can generate false knowledge that overwhelms the truth over a large space for a long time. What remains is a process that is always creative, always conjectural and always contextual.

  • No rule caps how large or how long-lived an error can be
  • False knowledge can outcompete truth across large regions for long periods
  • There is no shortcut that substitutes for conjecture and criticism
  • Knowledge is creative, conjectural and contextual, giving infinite room to improve

even a single false idea can create false knowledge that overwhelms the truth for quite a while in a large amount of space

Naval Ravikant · 33:30

it's always creative it's always conjectural it's always contextual which gives an Infinity of improvement ahead of us

David Deutsch · 33:30
#error-correction#progress#optimism#epistemology