Dawkins's Middle World claim says brains evolved on the savannah to survive, so we should expect to be cognitively closed to the very small, very fast, and very large. Deutsch rejects it on the grounds of universality. We already understand quasars and subatomic particles, and if the argument held, progress in mathematics and physics would have stopped long ago.

Middle World predicts we cannot agree on what quantum theory means because we are closed to the small
Universality of explanation is the reason the argument fails
Quasars and subatomic particles are already understood despite being outside Middle World
A valid version of the argument would have halted progress long ago
The claim is used to excuse the absence of consensus in quantum foundations

“this is the idea that Richard Dawkins has of middle World”
05:30

“this is flawed because of universality among other things we understand quazar in subatomic particles”
06:00

#universality#epistemology#quantum-foundations#dawkins

Hot Take· 5

Hot Take18:30

Deutsch: Constructor Theory might end up a curiosity, and that's fine

Pressed on predictions and tests, Deutsch offers only the lower bound on energy, and concedes it is weak because everyone believed it already. He expects the payoff to come when people formulate theories about new things that combine existing theories unexpectedly. Then he says outright that he cannot foretell the growth of knowledge and does not know whether the programme will work.

The energy lower bound falls out of Constructor Theory instead of being imposed as an extra condition
Deutsch dismisses the quantum theory of life as a blind alley
He thought the Multiverse solved free will in The Fabric of Reality and now thinks it does not
What the Multiverse does provide is a framework where counterfactuals carry more meaning
Rival theorists who promise to solve time or the Big Bang attract far more funding and attention

“I can't foretell the growth of knowledge I don't even know that construct theory is going to work”
20:00

“you just said there as a good poyan would that Constructor Theory could be completely wrong and it will just be a footnote in the…”
20:30

#constructor-theory#popper#free-will#prediction

Hot Take35:30

Athens is the counterexample to Matt Ridley's population threshold

Ridley argues you need a certain population size before progress can take hold. Deutsch is unconvinced and offers the earliest open societies as a counterexample. Athens was tiny by modern standards and a minuscule fraction of the world's population, yet produced an enormous burst of creativity that could have reached our level if it had lasted three hundred years.

Earlier enlightenments lasted a generation or two before ending
The current Enlightenment has run somewhere between three hundred and five hundred years
Athens was destroyed for reasons historians still work on, not from lack of capability
Previous enlightenments failed because they could not solve the problems that arose, and were too soon
Bad actors winning through force of numbers is treated as an irrelevant cause, not an intrinsic one

“I'm not sure that's true my counter example is the earliest open societies”
36:00

“our Enlightenment having lasted depends what you count but lasted Maybe 500 years maybe 300”
36:30

#enlightenment#progress#history#athens

Hot Take44:30

Free speech isn't much impaired; compelled speech is the real problem

Asked what the antidote is, Deutsch declines to say free speech must be protected at all costs, because he does not think it is currently much impaired. His distinction is between things you may not say and things you must say, with the latter arguably worse. The medieval guild is his contrast case: there, permission was required to make anything at all.

In a guild there was one sanctioned way to make a thing and no scope to invent another
Nobody today needs permission to speak; public speech merely carries constraints
The language remains universal even under those constraints
The threshold would be compelled speech becoming a substantial proportion of everything said
Soviet mathematicians still did mathematics while prefacing papers with praise for Stalin

“I could say freedom of speech is the thing that needs to be protected at all costs but I don't think it's true that freedom…”
44:30

“it's not that anybody that wants to say anything has to get permission it's that anyone who says things in public has to obey certain…”
45:30

#free-speech#compelled-speech#censorship#universality

Hot Take1:13:30

Woke is a fad, not a theory, because the negation of an explanation isn't one

Deutsch contrasts today's ideology with the communists of the twenties and thirties, who genuinely believed a theory about capital stealing labour's product. He argues the current version is a denial of good knowledge rather than a rival worldview, and offers a diagnostic: ask an activist to explain the causal chain their slogan implies and see whether an answer comes back.

A 1950 communist could explain how inflation follows from factory owners' behaviour; Deutsch doubts a modern activist can
The negation of an explanation is not an explanation
What remains is a rule about what must and must not be said, which is not a worldview
He classes it as a fad rather than a blind alley culture has evolved down
Because it is shallow, replacing the curriculum content might work without much resistance

“as I've often said the negation of an explanation is not an explanation”
1:14:00

“a communist in 1950 could have told you but I don't think a woke person”
1:15:30

#ideology#explanation#politics#curriculum

Hot Take1:17:30

Radicalisation is Britain's failure, and mass immigration is still the answer

Deutsch argues that the radicalisation of the children of Muslim immigrants, rather than the immigrants themselves, is the fault of Britain, multiculturalism, and an education system that trashes the host culture. His remedy is integration under one law rather than leniency. He then goes further, calling mass immigration the obvious solution to the population problem, conditional on screening for people who hate the country.

A century ago hundreds of thousands of Muslims served in the British army
Albania saved almost all its Jews from the Nazis, a counterexample to sweeping narratives
The failure was treating newcomers too leniently rather than too harshly
The remedy is that the same law applies to everyone
Applicants could be required to declare they love the country, and deported for lying on the form if they did not

“their children are radicalized whose fault is that that's not the fault of Islam that's the fault of Britain”
1:18:00

“the obvious way of solving the population problem is by immigration Mass immigration is the way to go”
1:24:00

#immigration#integration#britain#multiculturalism

Explainer· 7

Explainer01:30

Why emergent theories deserve the same status as microscopic ones

Deutsch rejects the habit, inherited from Newton, of shoehorning every physical theory into a bottom-up mould. He argues there is no legitimate criterion for excluding a class of theories from knowledge other than whether they are good explanations. Reductionism and holism should both be discarded as camps, because every level of emergence admits an explanation you actually need at that level.

The bottom-up mould creates the foundational problems in thermodynamics and statistical mechanics
An exact second law cannot be derived from time-reversible fundamental theories
Good explanation is the only admissible criterion for keeping a theory
You would never derive human evolution from particle collisions since the Big Bang
A theory deeper than the current paradigm will have to put emergent and microscopic theories on one level

“there cannot be a Criterion for excluding a set of theories from the body of knowledge”
02:30

“if you're trying to figure out how a steam engine works you're not going to do statistical mechanics and Trace every Collision”
03:30

#physics#reductionism#emergence#thermodynamics#explanation

Explainer04:30

What fundamental and real actually mean in Deutsch's worldview

Physicists hear fundamental and think reductionism. Deutsch uses it differently: fundamental ideas are the ones needed across more types of explanation, which is why he can call people a fundamental feature of reality. The parallel rule for what is real is whether the thing has to exist in your explanation, which he defends by asking in what sense laws of physics are real.

Fundamental ideas are those required by the widest range of explanations
There will never be an ultimate fundamental theory, because that would mean no problems remain
Something is real if the explanation cannot work without it
The rule for deciding something is real is not the same as a definition of real
Nobody can trip over a law of physics, yet nothing in physics is understandable without them

“fundamental ideas are the ones that are needed more types of explanations”
04:30

“in what sense are laws of physics real because we can't trip over one”
05:30

#epistemology#realism#physics#explanation

Explainer06:30

Knowledge as a crystal in the Multiverse

Deutsch's favourite link between epistemology and quantum physics. Knowledge is information that, once present, keeps itself in existence through error correction. Because there are many possible errors for every truth, universes that started unlike each other become alike in exactly the regions where error correction is happening, forming a crystal of similarity across the Multiverse.

Knowledge persists by error correction, not by being written down
Every truth has many more possible errors than instantiations
Error correction makes unlike universes converge locally
The crystal is the visible shape of that convergence across the Multiverse
Deutsch hopes to reformulate this inside Constructor Theory

“the idea of a crystal in the Multiverse if knowledge is kind of information which once it's there keeps itself in existence”
06:30

“in the Multiverse if there's error correction then there are a bunch of unlike universes which become alike in the region in which they're correcting…”
07:00

#multiverse#epistemology#error-correction#quantum

Explainer10:30

What problem Constructor Theory was built to solve

Deutsch explains the origin of Constructor Theory as a research programme whose defining principles are still unknown and will be found through applications. His first paper was philosophical, published in a philosophy journal, and set out motivations rather than results. He intended a quick paper and ended up with eighteen distinct motivations.

There is no good theory of initial conditions, yet no theory of final conditions is needed
Computation and information cannot be expressed in the initial-conditions-plus-laws-of-motion format
The thermodynamics problem is a third motivation
The change in worldview is simple to state but the ramifications are large
Chiara Marletto joined after catching a self-contradiction in one of his talks

“we don't have a good theory of initial conditions think carefully why should we need initial conditions when we are perfectly fine to have no…”
11:00

“I thought I would write a a quick paper and I I ended up with 18 motivations”
11:30

#constructor-theory#physics#research-programme

Explainer14:00

The lineage of timeless physics, from Wheeler-DeWitt to Constructor Theory

Deutsch traces the idea that time is emergent. Wheeler and DeWitt found a quantum gravity theory containing no time, only three-dimensional spaces with amplitudes, where time emerges from entanglement. Page and Wootters generalised it beyond gravity, and Deutsch's current work with Chiara Marletto aims to lay down rules such that any theory obeying them has time.

In the Wheeler-DeWitt picture nothing ever changes; time is how the spaces are entangled
Julian Barbour wrote a whole book making the case for a timeless universe
Page and Wootters extended it to any quantum theory, not just quantum gravity
Deutsch's reaction on seeing it was that this is what time is
Marletto already did the equivalent for probability, deriving it from rules rather than a specific theory

“they found a theory of quantum gravity in which there's no time time doesn't appear in the theory”
14:30

“so that static Universe still has time because of entanglement”
15:30

#time#quantum-gravity#constructor-theory#entanglement

Explainer29:30

Convergent evolution is a much stronger effect across the Multiverse

Asked whether error correction produces similar knowledge within a single universe as well as across the Multiverse, Deutsch says yes but weaker. Within one universe we see convergent evolution of body plans and of ideas, where people who never heard of each other reach the same conclusion. In the Multiverse every variant of a given size happens and gets error corrected, which makes convergence systematic.

Knowledge is that which tends to get replicated in the environment
Convergent biological evolution produces tiger-like animals independently in Australia
Independent simultaneous discovery is the memetic version of the same effect
Across the Multiverse all variants involving the same number of mutations occur, so convergence is systematic rather than accidental
Deutsch declines to draw a confident conclusion about how common life is

“we know that there's a convergent evolution of ideas because uh people can have an idea and person who they've never heard of has the…”
30:00

“convergent evolution is is a much stronger effect in the Multiverse”
30:30

#multiverse#evolution#error-correction#knowledge

Explainer32:30

Species and civilisations go extinct from too much fidelity, not too little

Deutsch corrects a common view that mutation rates are tuned to an optimum. All evolution inside the cell pushes toward greater fidelity, and no mutation reduces it. The consequence is that the vast majority of species that ever existed went extinct because they could not evolve fast enough to meet changed conditions. He then applies the same logic to civilisations.

Biology optimises the genetic code for fidelity, never for a useful mutation rate
Dawkins refuted the optimum-mutation-rate view cogently
Species die from error correction that is too strong for changing circumstances
The peacock's tail keeps growing even though it will eventually end the species
The overwhelming majority of civilisations died because memetic evolution was too slow

“species go extinct all the time for this very reason that they have too strong error correction not enough Evolution happens to cope with the…”
33:30

“the overwhelming majority of civilizations have gone extinct because the mimetic evolution hasn't happened fast enough”
41:30

#evolution#extinction#error-correction#memes

Story· 2

Story23:00

The quantum computer began as an experiment to test Everett

Deutsch recounts that he was not trying to invent a new mode of computation. The consensus said Everett's Multiverse made identical predictions to collapse theories; he thought that must be false and set out to design a test. Getting inside the observer required a computer that could not decohere, so he added quantum operations to a classical machine, and only much later saw what it was.

The paper was rejected as philosophy, and he set it aside for years
He told Penrose that if quantum theory is universally true then Everett is right, and Penrose replied that quantum theory isn't true
The test needs an observer that observes without collapsing, so it needs both AGI and quantum computing
The observer performs only about half a dozen quantum operations; the rest is classical thinking on coherent qubits
Charlie Bennett's line that the hardware is physics prompted him to recast Turing's analysis in quantum physics

“when I first came up with the idea of quantum quantum computer I didn't even think of it as a Quantum”
23:30

“I first thought of the experiment to test Everett and then much later thought about quantum computers”
26:30

#quantum-computing#everett#many-worlds#origin-story

Story1:04:30

The slave children who risked whipping to learn to read

Deutsch retells a favourite passage from Frederick Douglass's autobiography. Teaching slave children to read was illegal, so a woman taught them clandestinely at her house in the evenings, at risk of prison for her and a whipping for them. The boys came after a full day of forced field labour. Deutsch draws the contrast with how their descendants are described today.

The illegality attached to teaching, not to learning
The boys were already physically exhausted from the fields when they went
Douglass escaped north and made a great success
Deutsch says some of their peers stayed home, and that this was a legitimate call only they could make
His conclusion is that this is how humans are, not a special case

“slave children learning to read was illegal not for the slaves it was illegal to teach them to read”
1:04:30

“there were these teenage boys who were willing to risk a whipping and to uh go to these lessons”
1:05:30

#education#motivation#history#frederick-douglass

Frameworks in this episode

Anti-Rational Meme Detection

Layered Error Correction

Problem-First Learning

The What-Is-Your-Problem Test

Insights & moments

⚡Myth Buster· 1

The Middle World argument for human cognitive limits is wrong

◆Hot Take· 5