AMA | November 2022
What this covers
This November 2022 Ask Me Anything episode draws on Patreon-submitted questions, which Carroll selects and groups by topic. The conversation ranges widely across cosmology, quantum mechanics, philosophy of science, and ethics, with Carroll treating each domain through the lens of his broader view: that the universe admits a complete fundamental description, but that Emergent Description at higher levels—temperature, evolutionary processes, even the "worlds" of quantum mechanics—are real and useful even if not fundamental. The format allows him to address misconceptions, explain technical ideas plainly, and clarify where his own positions diverge from mainstream consensus.
The material covers distinct territories. Cosmologically, Carroll defends the idea that the early universe had low entropy and explains why claims about quantum fluctuations creating universes from nothing rest on confused notions of time. In quantum mechanics, he defends the Many-Worlds Interpretation against objections about why it remains heterodox, and works through technical details of wave function branching and the Born rule with Charles Sebens. On epistemology and method, he argues that intellectual maturity means admitting ignorance rather than forcing answers, criticizes how peer review favors safe over innovative work, and pushes back on the notion that materialism requires a "burden of proof" in some legal sense. Ethically, he examines tensions between empathy and rationality, defends deliberative voting systems against lottery schemes, and identifies the Golden Rule and Categorical Imperative as recurring cross-cultural values. Throughout, Carroll treats Poetic Naturalism as the orientation permitting multiple valid descriptions of reality without requiring them to collapse into a single level.
Carroll argues that the universe has a definitive description in principle, that emergent higher-level descriptions are real but optional human conveniences, and that intellectual honesty means embracing complexity and uncertainty rather than seeking simplistic, all-answering frameworks.
- There is a theory of everything because the universe exists and does something describable, and the brain being part of it is no obstacle to discovering it
- Many-worlds and other emergent descriptions (temperature, evolution, branches) are real but approximate handles on a fundamental reality
- Being an intellectual requires fallibilism and sincere truth-seeking, not having an answer to every question
Conservation laws arise from symmetries, and conservation fails where symmetry breaks—as expanding spacetime breaks time-translation symmetry, destroying energy conservation.
- Each conservation law corresponds to a symmetry, so conservation can fail where the symmetry is broken: conservation of momentum comes from spatial translational invariance, so a ball on a hill (a non-translationally-invariant environment) does not conserve momentum, just as expanding spacetime breaks the time symmetry behind energy conservation and the Higgs field breaks electroweak charge conservation.
“conservation of momentum comes from the fact that the universe is spatially translationally invariant. It's the same from place to place.”
- A study showing kids of all social positions—popular kids, bullies, and outcasts alike—all identified with Dumbo, the bullied outcast, when watching the film, illustrates that nearly everyone perceives themselves as the one being picked on, which is why there is essentially no genuine pro-bullying constituency.
“the answer is, everyone thinks they're Dumbo. Everyone thinks that they're the ones at the end of the day who are picked on”
A revolutionary physics breakthrough comparable to relativity or quantum mechanics is unlikely soon because today's puzzles like dark matter fit easily within the current paradigm, unlike the incompatibilities that drove earlier revolutions.
- A revolution comparable to relativity or quantum mechanics is unlikely in Carroll's lifetime because science progresses by explaining observational anomalies, and unlike circa 1900 (atomic structure, black body radiation strictly incompatible with classical mechanics), today's puzzles like dark matter are easily explainable within the current relativity-plus-quantum-mechanics paradigm.
“usually the way that physics or other sciences actually make progress is by trying to explain observational or experimental anomalies, and right now in fundamental physics, we don't know of that many such anomalies”
Entanglement requires a quantum system in superposition interacting with another such that different superposition components interact differently with it.
- Entanglement occurs not merely from interaction or proximity but when at least one quantum system is in a superposition and it interacts with another system such that the different parts of the superposition interact differently; a spinning particle dropped in Earth's gravity interacts but does not entangle because both spin components evolve identically, whereas a photon absorbed by only the awake half of a superposed cat creates entanglement.
“entanglement happens when that system interacts with some other system in such a way that the different parts of the superposition interacts differently”
General relativity itself is not time-asymmetric; rather, the expanding-universe solution describing our cosmos is time-asymmetric, and energy fails to be conserved because spacetime's metric is non-stationary.
- General relativity is not necessarily time-asymmetric; rather, the specific cosmological solution describing our expanding universe is time-asymmetric, and energy is not conserved because the spacetime metric is non-stationary—if spacetime were perfectly stationary, energy would be conserved, and there is a definite covariant equation relating spacetime's behavior to how energy changes.
“it's not that general relativity is necessarily time asymmetric, it's that the specific solution to the equations of general relativity corresponding to our real world in cosmology is time asymmetric”
Position and momentum are symmetric in abstract mechanics, but the real-world Hamiltonians have asymmetry: particles interact more strongly when nearby in position than when they have similar momenta.
- While position and momentum are perfectly symmetric in abstract Hamiltonian mechanics where any Hamiltonian is allowed, the specific Hamiltonians describing the real world have an asymmetry: particles interact more strongly when nearby in position (bumping or via electromagnetic/gravitational effects) but there is no principle that particles with similar momenta interact strongly, and this difference traces to position and momentum appearing differently in the real-world Hamiltonian.
“particles will interact with each other, more strongly at least, when they are nearby in position... There's no principle in the real world that says, if two particles have the same momentum, they will interact in some strong way.”
Pure rationality fails by attending to limited information, while empathy matters because imagining ourselves in others' shoes opens our eyes to information and concerns we would otherwise miss.
- Although rationality may seem a better moral guide than empathy because people empathize most with those like themselves, the failure mode of pure rationality is that we can feel rational while only attending to limited information; empathy matters because imagining ourselves in others' shoes opens our eyes to information and concerns we would otherwise ignore.
“it's really easy to think you're being rational when you're not”
A sign of intellectual honesty is the willingness to say 'I don't know' rather than always having an answer, distinguishing genuine truth-seekers from those who may have over-simplified.
- If you think you can answer every single possible question, that is not a sign you are super smart but a sign you may have gone wrong somewhere and are over-simplifying; honest truth-seekers must be able to say 'I don't know,' as illustrated by a New Age believer who deconverted upon noticing skeptics admitted ignorance while her New Age friends always had an answer.
“If you think you can answer every single possible question, that's not a sign, you're super smart, that's a sign you might have gone wrong somewhere.”
An equation can be complete and revolutionary before its detailed solutions are found, as Einstein's 1915 field equation was complete yet interesting exact solutions came later.
- An equation can be complete and revolutionary before its solutions are found: Einstein's 1915 field equation was complete and he immediately solved it approximately (enough to predict light deflection and Mercury's precession), but its interesting exact solutions came later when Schwarzschild found a specific metric G_mu_nu satisfying it—'solving' meaning finding the actual value of the unknown (the metric) that satisfies the equation, just as solving x²=4 yields x=±2.
“you can propose an equation and have a very good reason to believe that it is correct... without yet knowing what the solutions are”
We cannot demand the universe operate according to our preferences; the good news is that it permits us to figure it out piecemeal, regime by regime.
- We don't get to make demands of the universe about how it should operate—it would be nice but no requirement for a fundamental theory to uniquely predict our world, and Carroll is happy if the fundamental theory predicts many possibilities with us as a statistical prediction; the non-trivial good news is that the universe operates such that we can figure it out piecemeal, getting better regime by regime.
“we don't get to make demands of the universe and how it operates. I wanna figure out how the universe operates, I don't wanna tell it how to operate.”
AI tools should be used like Wikipedia—as pointers to verify and check independently—rather than trusted as definitive answers.
- AI tools should be treated like Wikipedia—useful as a pointer to find things or to lay out a calculation you can independently check, but not trusted to be the definitive answer—and seeking out mathematical tools for physics ideas (as Einstein did with Riemannian geometry) is a common and ongoing practice.
“I think AI is exactly the same way in its current incarnations, I would not trust it to do a calculation, but if it laid out a calculation, I could certainly check it.”
Podcast episode selection is based on whether there is something interesting to discuss and learn, not on whether a topic is the most urgent issue facing humanity.
- There is zero connection between what Carroll thinks is the most urgent issue facing humanity and what he covers on the podcast; the criterion is whether there is an interesting thing to discuss and something to learn, not importance—he doesn't believe in the therapeutic value of complaining and refuses to do an episode merely to reassure listeners they are virtuous about a topic like climate change they already accept.
“there's zero connection between what I think is the most urgent issue facing humanity and what I need to have on the podcast”
The James Webb Space Telescope is extremely unlikely to alter fundamental physics credence because it was not designed to probe fundamental laws but rather galaxy formation and exoplanet atmospheres.
- It is extremely unlikely that anything discovered by the James Webb Space Telescope will alter credence on physics fundamentals, because the telescope was not designed to probe fundamental laws—it looks at early galaxy formation and exoplanet atmospheres—even though looking at the universe in new ways can always produce surprises.
“none of these are gonna change physics fundamentals”
Human progress is not doomed because entropy only increases in closed systems while human systems are dramatically open, receiving low-entropy solar radiation and returning high-entropy radiation.
- Human progress is not doomed to merely shift disorder around, because entropy only increases in closed systems and human systems are dramatically not closed—the biosphere receives low-entropy radiation from the Sun and returns high-entropy radiation—so while universal entropy rises, local entropy stays low, and the relaxation timescales of large-scale systems vastly exceed a human lifetime, making the detailed dynamics matter far more than the bare statement that entropy increases.
“entropy goes up in closed systems, which the usual way that we think of human systems, they're not closed”
Treating the universe as a thermodynamically closed system is justified because there is no theoretical or experimental reason it shouldn't be, and it isn't clear what an open universe would mean.
- Treating the universe as a thermodynamically closed system is justified because there is no theoretical or experimental reason it shouldn't be, and it isn't even clear what an open universe would mean—an open system exchanges energy/information with another system, but any such other system could simply be redefined as part of the universe—so the simplest assumption compatible with everything we know is to treat the universe as closed.
“if our universe was exchanging energy information with another system, I would just redefine universe to include that other system too”
A genuine intellectual distinguishes themselves by fallibilism—recognizing and fixing mistakes—and by being sincere, open-minded, careful, and honest in seeking truth.
- What distinguishes a genuine intellectual is not merely seeking truth (which conspiracy theorists also believe they do) but being fallibilistic—capable of recognizing and fixing mistakes—and being sincere, open-minded, careful, and honest in the search; one can either restrict the definition this way or simply admit conspiracy theorists are intellectuals who are just very bad at it.
“you would want them to be fallibilistic, you would want them to be capable of understanding when they've made a mistake and therefore fixing the mistake”
A sufficiently large quantitative jump in intelligence may be indistinguishable from a qualitative jump, and there may be no qualitatively different form of intelligence above humans.
- Carroll is not sure there is a qualitative jump from cat to human intelligence, suspecting that a sufficiently large, obvious quantitative jump is indistinguishable from a qualitative one; he suspects there is no qualitatively different form of intelligence above humans, but allows that something could be so much more capable—more logical, rational, mathematical—that it would seem like a completely different kind of thing from our point of view.
“a really really, really noticeable, big, obvious quantitative jump is indistinguishable from a qualitative jump”
It would not be ethical in the real world to assassinate an elected leader because adopting such a principle would license many wrongful assassinations, and only in abstract certainty of world-ending harm would it be permissible.
- In the real world it would not be ethical to assassinate even a destructive elected leader, because adopting a principle that elected officials you expect to do terrible things can be assassinated would license many wrongful assassinations—the issue is not who deserves to live but who gets to decide—and you could be wrong, which is why we have institutions; only in abstract philosophy-land with certainty of world-ending harm would it be permissible.
“if you have a principle that says that elected officials who you think are gonna do terrible things can be assassinated, then a lot of people are gonna take that principle and run with it”
Modern physics does not use Aristotelian causes as fundamental to dynamical evolution; rather, causes are higher-level emergent phenomena rather than fundamental features of reality.
- Modern physics does not use Aristotelian notions of cause and effect, and the key distinction—elaborated in Carroll's book The Big Picture—is between the tradition (Aristotle through Leibniz's principle of sufficient reason) that treats causes as fundamental and front-and-center in the dynamical evolution and explanation of the world, versus the modern view that causes are a higher-level emergent phenomenon rather than fundamental.
“What's important here is the idea that causes are fundamental in the universe rather than higher level emerging phenomenon.”
Statistics, probability, and Bayes's Theorem should be taught much more widely so everyone understands them by the time they leave high school.
- Beyond basic algebra, trigonometry, geometry, and the concepts of derivative and integral, statistics and probability should be taught much more—including the axioms of probability, marginal and conditional probabilities, and Bayes's Theorem—so that everyone understands Bayes's Theorem by the time they leave high school.
“I think that everyone should understand Bayes's Theorem when they come out of high school”
Strict materialism does not assume a burden of proof because burdens of proof are legal concepts unsuitable for truth-seeking, though positing a deity is one extra assumption requiring justification.
- A strict materialist who rejects any higher organizing principle does not assume a burden of proof, because burdens of proof are a legal concept and using legal proceedings as a paradigm for finding truth is a very bad way to find truth—when people mutually seek truth they should work together with no one bearing a burden of proof, though positing a deity is one extra assumption that requires a reason.
“Using legal proceedings as a paradigm for finding the truth is a very, very bad way of finding the truth”
There is definitely a theory of everything because the universe exists and does something, and that has a description; the human brain being part of it poses no obstacle to discovering it.
- There is definitely a theory of everything because the universe exists and does something, and whatever that is has a description; furthermore the human brain being part of that theory is no obstacle to discovering it, because a theory of everything is a complete specification of the patterns and laws the world obeys, not a complete precise specification of the world itself.
“a theory of everything is not... supposed to be a complete and precise specification of the world, it's supposed to be a complete specification of the patterns, the laws that the world obeys”
The current historical moment is genuinely crucial because things are changing rapidly and humanity now has far more leverage over the whole globe than ever before.
- We are justified in thinking of the current point in time as historically crucial because things are changing rapidly and humanity now has far more leverage over the whole globe—through nuclear weapons, climate change, and biological hazards—than it ever had before, even if a point 100 years from now may seem more crucial still.
“we can do tremendous damage to the planet, whether it's nuclear war or climate change or potential biological hazards one way or another”
The phrase 'universe created as a random quantum fluctuation' is not sensible because a fluctuation requires something to fluctuate from, whereas a universe with a beginning just means a first moment of time.
- The phrase 'the world was created as a random quantum fluctuation' is not a sensible idea on the market, because a fluctuation requires fluctuating from something to something else, whereas if the universe literally had a beginning that just means there was a first moment of time—not that it fluctuated out of pre-existing nothingness; there was no time before the first moment to ask how long it was non-existent.
“if it's literally the creation of the world, then there was nothing pre-existing for it to fluctuate from”
Poetic naturalism could plausibly apply to ethics, with a fundamental but complicated moral theory underlying practical rules of thumb and approximations.
- Poetic naturalism could plausibly be applied to ethics: there could be a fundamental, universal (though possibly constructed, not objective) moral theory so wickedly complicated in practice that we instead use rules of thumb and approximations at higher levels—analogous to how the Standard Model underlies but isn't used for engineering a building—and probably something like that is what we actually aspire to do.
“is there some fundamental moral theory, and it doesn't need to be objective, it can be something constructive... but it's supposed to apply to all situations”
The Standard Model is not merely fitted to each experimental result; the number of experiments vastly exceeds parameters, making the theory highly over-constrained and remarkably resilient.
- The claim that the Standard Model merely agrees with experiment because parameters are tuned to each result is completely false; the number of experiments vastly exceeds the number of parameters, so once a parameter like the Higgs mass is fixed by one measurement, all decay rates become predictable with no extra free parameters, making the theory highly over-constrained and remarkably resilient.
“It is highly highly over-constrained and remarkably resilient at agreeing with the experiment.”
In Everett interpretation, everyone agrees the wave function evolves by the Schrödinger equation; dividing it into separate 'worlds' is a human-convenient, approximate, emergent description.
- In the Everett interpretation the only thing everyone must agree on is that there is a wave function evolving entirely by the Schrödinger equation; dividing that quantum state into separate 'worlds' is a human-convenient, approximate, emergent description—just like describing a gas by temperature versus tracking individual molecules—so people like Carroll and Deutsch can legitimately disagree about the best way to divide it before decoherence.
“how we take the quantum state of the universe and divide it up into worlds is a human convenience”
Democracy is under threat partly because safeguarding it is genuinely unpopular as a priority, and many prefer one competent leader to the messy voting process, requiring deliberate education to sustain.
- Democracy is under threat partly because it is genuinely not that popular—safeguarding democracy polls near the bottom of priorities for both parties, and many people would rather have one competent person in charge than the messy process of voting—so its long survival is almost a miracle that requires deliberate education to sustain.
“Deep down democracy is not that popular, and it's almost a miracle has lasted this long.”
Tying health insurance to employment is bizarre and a major reason the US has similar healthcare outcomes to other developed countries while spending roughly twice as much per person.
- Tying health insurance to employment status is bizarre and is a big part of why the United States has similar healthcare outcomes to other developed countries while spending roughly twice as much money per person, making the system wildly inefficient and obviously improvable without great difficulty.
“it is a big part of why the United States has similar healthcare outcomes to other developed countries, but spends twice as much money per person”
There is no clean list of core values enabling societies to thrive, though a recurring cross-cultural value is treating others as having inherent dignity, captured by the golden rule and Kant's categorical imperative.
- There is no clean list of 'core values' that enable societies to thrive because values like justice and fairness overlap, but a recurring cross-cultural value is treating other people as having inherent dignity—captured by versions of the golden rule and Kant's categorical imperative—which sometimes constrains simplistic utilitarian greatest-good reasoning.
“we should act as if there is some inherent dignity in other human beings”
In debate, the goal should be to be understood and convey why naturalism is fulfilling, not to win points in the stylized competitive sense that has nothing to do with searching for truth.
- In debating William Lane Craig, Carroll's goal was not to win in the stylized point-by-point sense of competitive debate (which has nothing to do with searching for truth) but to be understood—to help the audience grasp why someone could be a rational and even good atheist and why naturalism is a fulfilling worldview—so he used his final speech to give that sales pitch rather than rebut Craig's last speech.
“I'm not interested in winning the debate in the sense of a high school or college debate competition”
Referees are more likely to reject imaginative non-mainstream papers because they dislike brand-new ideas, while influential papers were more likely rejected on first submission than dull papers.
- In physics, more imaginative or non-mainstream papers are more likely to get rejected because referees dislike brand-new ideas, while boring run-of-the-mill papers that play by the rules get accepted—and studies show highly influential papers were more likely to have been rejected on first submission than dull papers, which is something to keep in mind about the refereeing system.
“highly, highly influential papers are more likely to have been rejected the very first time they went to a journal than boring, dull papers”
Classical general relativity should be extremely good at describing a medium-sized black hole; the most valuable experiments would test for deviations from predicted gravitational fields.
- Black holes are not actually that mysterious in how they behave, because classical general relativity should be extremely good at describing a medium-sized black hole; the most valuable experiments near one would test for deviations from GR's predicted gravitational field and, for a small black hole, observe Hawking radiation, with any deviation being revolutionary but unlikely.
“good old Einstein's classical general relativity should be super good at explaining what it is like. We should have already a pretty good picture”
Biological systems like the brain are enormously powerful but not well-designed due to random variation and shifting environments, so transcendent experiences are unsurprising and may serve evolutionary purposes or arise by chance.
- Because evolution is not a straight-line design but works through random variations in the DNA blueprint playing out in shifting environments, biological systems like the brain end up enormously powerful but not well-designed—with redundancy in some places, fragility in others, and inexplicable wiring—so transcendent or mystical aesthetic experiences are unsurprising and may have served a good evolutionary purpose or arisen partly by chance.
“we have to work as previously mentioned, with this framework of looking at random fluctuations in the underlying blueprint from the DNA and seeing how they play out in the real world, which is a very hit-or miss way of doing things”
Many-worlds is not consensus because its metaphysics radically differs from everyday experience, and though simple to state, much work is needed to connect it to observations, risking self-deception.
- There are decent (though Carroll thinks ultimately wrong) reasons many-worlds is not the consensus view: its underlying metaphysics is radically different from everyday experience, and although the theory is simple to state (a vector in Hilbert space evolving by the Schrödinger equation), much work is needed to connect it to observations, with an obvious risk of fooling oneself—as he candidly admits about deriving the already-known Born rule—whereas hidden-variable and objective-collapse theories map more directly onto observed particles or collapses.
“the metaphysics, the underlying view of the world presented by many worlds, is radically different than the view of the world that we're familiar with from our everyday experience”
Particle spin is not a spinning ball but genuinely a field with angular momentum; because there are no particles, only quantum field descriptions, spin's angular momentum is real field angular momentum.
- The spin of an elementary particle is not a little spinning ball but is genuinely a field with angular momentum; because there are no particles—only convenient descriptions of measured quantum fields—the angular momentum in particle spin is real field angular momentum that contributes to the conserved total angular momentum of larger systems.
“it's not that there's a little dot or a little baseball that is spinning, it's the field that has angular momentum”
We cannot directly test that the early universe had low entropy because we observe photons in telescopes, making the inference back to low entropy the 'Past Hypothesis,' a self-consistent assumption never an experiment.
- We cannot directly test that the early universe had low entropy because we don't observe the early universe—we observe photons landing in our telescopes—and the inference back to a low-entropy past assumes entropy has been increasing the whole way; hence it is the 'Past Hypothesis,' a self-consistent assumption compatible with what we see, but never an experiment that can rule out a high-entropy universe that randomly fluctuated into its current state.
“it assumes the past hypothesis because we don't actually see the early universe, we see photons landing at our telescope”
Emergent complexity generically requires two special features: a hierarchy of mass scales enabling energy dissipation and a special low-entropy initial state; without these we would expect chaotic mess.
- Generically, most cellular-automaton rules and most physical systems would not produce simple coherent large-scale emergent structures; our real-world emergent complexity relies on two special features—a hierarchy of mass scales (including massless photons that allow dissipation by carrying energy away) and the past hypothesis of a special low-entropy initial state—without which one would expect chaotic mess rather than emergence.
“without those two very special features, I think that generically, you would not expect simple, coherent large-scale structures”
Lottery voting is undesirable because it would let a small crazy faction win proportionally often, whereas smoothing out extremes is an intentional virtue of voting systems ensuring most aren't dissatisfied.
- Lottery voting is undesirable because, although it is proportionate and gives voters more power, it would let a 10% crazy/extremist faction win 10% of the time, whereas smoothing out extremes is actually an intentional virtue of voting systems—democracy should ensure most people aren't too dissatisfied rather than literally representing every fringe, so the slow decay of a voter's influence in plurality systems is a feature, not a bug.
“in a lottery voting system, the crazy fringes would win 10% of the time and 10% things happen a lot”
Carroll is more open-minded about whether the universe is finite or infinite; infinity solves arrow-of-time problems, though a finite universe could in principle host a well-posed arrow of time.
- Carroll is more open-minded than he used to be about whether the universe is finite or infinite: an infinite universe that lasts forever solves arrow-of-time problems (avoiding Boltzmann-brain-style repetition), but a universe finite in both states and time—bounded or cyclic—could in principle host a well-posed arrow of time, though he doesn't know how to make that work.
“if the universe lasts forever then it better have an infinite number of things it can possibly do, because otherwise we'll just do some subset of things over and over and over again”
Truly learning a subject requires knowing techniques for solving equations through problems and exercises, which is deliberately omitted from The Biggest Ideas in the Universe to compress information.
- Truly learning a subject requires more than knowing the equations—you must know techniques for solving them by doing problems and exercises—which is precisely the part deliberately omitted from The Biggest Ideas in the Universe, allowing it to compress so much information by not tickling that part of the brain.
“if you truly want to learn something, you can't just know the equations, but you have to know techniques for solving them, for doing the problems, for doing the exercises”
Dark matter is unsurprising because it is simply a particle with substantial mass abundance that avoids strong or electromagnetic interactions, which is very easy to posit.
- It is not surprising there is dark matter, because dark matter is simply some particle with substantial mass abundance that doesn't interact via the strong or electromagnetic interactions—which is very easy to posit, as neutrinos already are dark matter particles (just not the right kind, being too low-mass and fast, making them hot rather than the cold dark matter that fits the data).
“all dark matter is, is some particle that has the substantial abundance of mass in the universe that doesn't interact either through the strong interactions or the electromagnetic interactions”
When physicists speak of 'different laws of physics' in the string landscape, they mean the same underlying quantum field theory and relativity but different low-energy physics from different extra-dimension configurations.
- When physicists speak of universes with 'different laws of physics,' in the string-theory landscape they mean the same underlying quantum field theory and relativity but different low-energy physics—different particles, forces, gauge symmetries, coupling strengths, cosmological constant values, even different numbers of spatial dimensions—arising from different ways the extra dimensions are curled up.
“those ways all give rise to different low-energy laws of physics, different local visible laws of physics”
LISA uses the same laser-interferometry principle as LIGO, detecting gravitational waves through slight variations in relative distances between perpendicular laser beams.
- LISA will work on the same principle as LIGO—laser interferometry that looks for slight variations in the relative distance traveled by perpendicular laser beams, detected as a disruption of their interference cancellation—so it does not require knowing the absolute distance, only changes in it caused by a passing gravitational wave.
“you don't even really need to know the actual distance that the light beam travels to a great precision because all you're looking for are slight variations of that distance”
Euclidean geometry dominated until hyperbolic geometry appeared—a uniform negatively curved surface not embeddable in three-dimensional Euclidean space yet having its own geometry.
- Euclidean geometry survived as the single solution for so long because a two-dimensional sphere is just a subset of three-dimensional Euclidean space requiring no new axioms; the breakthrough came with hyperbolic (negatively curved) geometry, since a uniform negatively curved two-dimensional surface cannot be embedded in three-dimensional Euclidean space yet still has its own geometry—a wholly new kind of thing.
“a uniform negatively curved two-dimensional surface, cannot be embedded in three-dimensional Euclidean space. But it still does have a geometry of its own.”
Evolution specifically requires information passed between generations with random variation; it would be natural on any life based on genetic information anywhere in the universe.
- Evolution is not tautological 'survival of the fittest'; it specifically requires information passed from one generation to the next with the possibility of slight random variation, so it doesn't count as evolution merely to have progeny—and because genetic information is never copied perfectly, evolution would be natural on any life based on genetic information anywhere in the universe.
“it only counts as evolution, if you have some information in one generation that is passed on with the possibility of slight variations to the next generation”
Recent research shows widely used Dynamic Stochastic General Equilibrium models are essentially nonsense, with parameter recovery bearing no relationship to inputs and arbitrary relabeling yielding equal fits.
- A recent paper by McDonald and Shales reportedly shows that widely used Dynamic Stochastic General Equilibrium economic models are essentially nonsense: feeding a model's own predicted history back into it recovers parameter values bearing no relationship to the inputs, and even swapping the labels of inflation and unemployment rates yields just as good a fit—illustrating the deep difficulty of modeling and forecasting complex social behavior.
“there was essentially no relationship between the values of the parameters that they derived and the ones they had put in”
We have theoretical grounds to lower credence in cosmologies predicting many Boltzmann brains because such scenarios are cognitively unstable and coherent alternatives fitting data exist.
- We have theoretical grounds, not just practical reasons, to lower credence in cosmological theories predicting many Boltzmann brains, because there is no principled justification for believing one theory true while acting as if another is true; the Boltzmann-brain scenario is cognitively unstable since you can draw no conclusions and know nothing about the world under it, and since coherent alternatives that fit all the data and assume a low-entropy past exist, we should advocate those instead.
“it's no way to go through life accepting that because you can draw no conclusions, you do not know anything about the world, if that's the cosmological scenario you accept”
Within many-worlds, if any probability rule emerges it must be the Born rule by Pythagoras; the real task is justifying why there is a probability measure at all.
- Within many-worlds, if the theory implies any probability rule at all, it is guaranteed to be the Born rule (probability proportional to amplitude squared) because of Pythagoras—the hypotenuse squared equals the sum of the squares of the sides; the real task is to justify why there is a probability measure at all, which Carroll and Sebens did via self-locating uncertainty when the wave function branches, so there is no separate 'mechanism' left to explain.
“it is crystal, 100% super-duper clear that if you think that the theory implies any probability rule at all, that probability rule is going to be the Born rule”
There should be much higher inheritance taxes, with government better spending the money by weighing all needs responsibly versus wealthy individuals' idiosyncratic philanthropic preferences.
- There should be inheritance taxes, probably much higher than current ones, with the money spent better by government afterward; the objection that government won't use the money well is easily defeated by comparing how government typically spends versus how wealthy individuals with quirky interests spend, since government has the duty to weigh all needs responsibly while individual philanthropy reflects idiosyncratic preferences.
“I do think that there should be inheritance taxes, and I think there probably should be a lot more than there are.”
The anti-Humean view of laws of nature is not literally a 'God view,' and plenty of atheists are anti-Humeans, so the phrase should not be taken literally.
- The anti-Humean view of laws of nature is not literally a 'God view'—that was Barry Loewer's cheeky terminology between fellow Humeans—and while there is a family resemblance between the idea that laws have a separate identity from the physical universe and the idea that God has a separate existence, plenty of atheists are anti-Humeans, so the phrase should not be taken literally.
“There's plenty of atheists who are anti-Humeans.”
A finite number of degrees of freedom does not imply a smallest non-zero probability because quantum states form a continuous spectrum from continuous complex amplitudes.
- A finite number of degrees of freedom in a volume does not imply a smallest non-zero probability of any event, because quantum mechanical states form a continuous spectrum: any superposition of basis states is a valid state, so a single particle (one degree of freedom, up or down) has infinitely many possible states, and probabilities come from continuous complex amplitudes, not from counting states.
“in Quantum Mechanics, states are continuous states are any superposition of any set of basis states”
History does not have a direction in any especially useful sense; patterns emerge that we simplify after the fact, but 'direction' wrongly implies inevitability or external control.
- History does not have a direction in any especially useful sense—like the evolution of the whole biosphere, patterns and phases emerge that we can simplify after the fact, but 'direction' wrongly implies either something directing it or inevitability, whereas an empirical trend like 'colonization declined' exerts no control: if a country wants a colony and no one stops them, they will get one regardless of historical trends.
“I don't think that history has a direction in any especially useful sense, no.”
Questions about counting branches in many-worlds are not well-posed because the number of branches is infinite, uncountable, and uninteresting—you can only integrate over branch sets.
- Questions about counting the number of branches in many-worlds are not well-posed because the number of branches is infinite—you can integrate over sets of branches to get a weight, but there is no fact about there being two more branches in a given period; more branches are morally created the longer you wait, regardless of whether you isolate the object, but the precise count is uncountable and uninteresting.
“If the number of branches is infinite, there's no way to count, you can integrate over certain sets of branches and get a weight that you can apply to them, but the total number is infinite”
The wave function of the universe can be viewed as a single unified thing, but categories of fundamentally one versus multiple don't fit what the universe really is.
- It can be scientifically valid to view the wave function of the universe as a single unified thing, but the categories of fundamentally one versus fundamentally multiple don't fit easily onto what the universe really is—just as a red block and a blue block are both one set and two blocks—so the universe is plausibly a single monistic wave function that is also usefully divisible into sub-factors (quantum mereology).
“the categories of being fundamentally one or being fundamentally multiple don't fit very easily onto the question of what the universe really is”
Black hole merger produces no real mass-to-gravitational-wave conversion; all energy in masses and orbital motion is contained in spacetime curvature smoothly evolving with some energy radiating away.
- When black holes merge there is no real conversion of mass to gravitational waves, because the mass is just energy (via E=mc²) and all the energy—in the holes' rest masses, their orbital kinetic energy, and the emitted waves—is contained in the curvature of spacetime; it is one configuration of curved spacetime evolving smoothly into another with the same total energy, some now radiating away as gravitational waves.
“all of the energy in the black holes and in the gravitational waves is all contained in the curvature of spacetime”
Many black holes in a line around a closed universe would not form a stable cylindrical black-tube horizon because you can increase entropy by breaking it into individual black holes.
- Putting many black holes in a line around a closed universe would not form a stable cylindrical 'black tube' horizon in our 3+1 dimensional spacetime, because you can increase the horizon area—and thus the entropy—by breaking it into individual black holes, which is what black holes tend to do; though such black strings can exist stably in other numbers of dimensions or universe sizes.
“the reason why it would not be stable is you can increase the area just by beating it off into individual black holes, and that increases the entropy”
Free speech is crucially important but not simple; drawing the line requires difficult nuanced thinking balancing free speech against other values rather than treating it as absolute.
- Free speech is crucially important but not simple—it is being used as a cudgel on all sides in ways that dumb down discourse—and because we all agree some speech should not be spread everywhere while other speech should be unfettered, drawing the line requires difficult, nuanced thinking that balances free speech against other values rather than treating it as an absolute.
“Free speech is crucially important, but it's also not simple, it's complicated”
The most biologically plausible way for a creature to breathe fire would be belching highly flammable gas in an ignition-capable environment or storing two separate stable gases combining outside the body.
- The most biologically plausible way for a creature to breathe fire would be to belch a highly flammable gas in an environment where sparks ignite it, which is safer than starting an internal fire, or even better to store two separately stable gases (like hydrogen and oxygen) that become flammable only when combined outside the body, as in rocket engines.
“what you would want is not to actually be able to breathe fire, but to belch some sort of highly flammable gas in an environment where things light on fire all the time by themselves”
Regarding public policy one should be conservative in the old-school sense: if a system works reasonably well, there is sensible reluctance to dramatically change it due to potential unintended consequences.
- When it comes to public policy one should be conservative in the old-school lowercase-c sense: if a system works pretty well, there is sensible reluctance to change it dramatically because a new system may have unintended consequences, so a tie goes to the present system—though this is not the reason people resist the obviously feasible US reforms.
“a tie goes to the present system in those kinds of circumstances”
Being well-off does not obligate stopping use of public goods like libraries; a good response is redirecting what you'd have spent to worthy charity while supporting public institutions.
- Being well-off does not create an obligation to stop using public goods like libraries; rather, if you can afford to give more than necessary, a good response is to redirect what you'd have spent to a worthy charity, and supporting public libraries (even via late fines) is good.
“I don't think you're under the obligation to not use the public library, but maybe the money you would have spent on a book, give to some charity”
In the holographic principle, the bulk has one more spatial dimension than the boundary while both share the same time dimension; time merely goes along for the ride.
- In the holographic principle (e.g. AdS/CFT), the bulk has one more dimension of space than the boundary while both have the same single dimension of time, so the reduction is from a number of spatial dimensions to one fewer spatial dimension—time just goes along for the ride and does not change.
“the bulk has one more dimension of space than the boundary does, but it has the same number of dimensions of time, namely one”
Twitter's fate under Elon Musk reminds us that important things don't last forever and we live nowhere near equilibrium where rapid change makes predicting the future extremely difficult.
- The current obsession over Twitter's fate under Elon Musk is a reminder that things which become very important don't last forever, and that we are living in a time in history nowhere near equilibrium where things change rapidly, making it very hard to predict the future or assume anything about the current state is typical or permanent.
“these things that become very important don't last forever, and that in turn is a reminder of the fact that we are living in a time in history where we're really nowhere near equilibrium”
Graham's number is a finite number whose digit storage would collapse into a black hole, unlike pi's infinite expansion which no one ever thought the brain could hold.
- While both pi and Graham's number can be defined in a finite number of bits, the difference Tony Padilla was making is that Graham's number is a finite number you might think you could hold in your head but cannot, because the brain's storage capacity needed to hold all its digits would collapse into a black hole—whereas no one ever thought pi's infinite decimal expansion could be held.
“the storage capacity of the brain that would be necessary to do that would collapse into a black hole, that is the implication”