What this covers
This is an hour-long question-and-answer session from Sean Carroll's Mindscape podcast in which Carroll responds to Patreon-submitted questions spanning physics, philosophy, and practical matters. Rather than a lecture, it reads as a working conversation: Carroll picks questions he judges worth his time and sometimes clusters related ones, moving between quantum mechanics and black holes, consciousness and free will, cosmology and taxation, chemistry and the demarcation of science.
The session works across distinct territory. On fundamental physics, Carroll reframes classical puzzles—why black holes are "bald," why the early universe was low-entropy—as questions about quantum field theory, entropy, and weak emergence. On quantum interpretation, he argues the Many-Worlds Interpretation solves the measurement problem via self-locating uncertainty, and that Bayesian Reasoning governs how to weigh competing theories. On philosophy, he uses coarse-graining and emergent description to dissolve debates about free will, human identity, and consciousness—showing why "could have done otherwise" is compatible with determinism, and why a physicalist cannot coherently imagine philosophical zombies. He tackles cosmology (the size of the universe at the Big Bang, expansion rates, dark matter and dark energy), the uncertainty principle, the demarcation problem via falsifiability, deity versus advanced aliens, and optimal approaches to social change, touching on Elizabeth Anderson and moral philosophy. Technical questions sit alongside practical ones about seafood toxins and fair taxation, with Carroll treating each at the level of seriousness its substance warrants.
Carroll argues that most deep puzzles in physics and philosophy dissolve once you take emergence and quantum mechanics seriously: macroscopic concepts (black hole baldness, free will, human identity, consciousness) are higher-level descriptions of a fundamentally quantum world, and conflating microscopic and emergent levels is the source of most confusion.
- Black holes are bald because they are maximum-entropy states, reframing a classical GR fact in quantum/thermodynamic terms
- Free will and 'could have done otherwise' are consistent with determinism because we lack microscopic initial conditions and must use coarse-grained descriptions
- Consciousness and zombies are not conceivable for a physicalist because consciousness is weakly emergent from physical behavior
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
Black holes have no hair because they are local maxima of entropy — the highest-entropy state you can fit into a region of fixed size — and maximum-entropy states are featureless given their macroscopic constraints (mass, charge, spin), just as a heated object in a sealed oven radiates black-body radiation independent of its material, color, or shape.
“Black holes have the highest entropy that you can fit into a region of fixed size. They are, in some sense, local maxima of entropy, and what that means is operationally that they're all gonna look like.”
Saying dark matter and dark energy exist IS saying the theory was wrong: the original theory (visible matter plus general relativity) made a prediction that came out wrong, so adding dark matter/energy is changing the theory just as much as changing the theory of gravity would be — and cosmologists have tried modifying gravity instead, but it doesn't fit the data, while dark matter and dark energy fit very well.
“when we say there's dark matter and dark energy it is precisely that the theory is wrong, that is what we're saying.”
Unlike most land animals we eat (cows, chickens, pigs, which are herbivores), the fish we eat are themselves carnivores eating other fish, so toxic chemicals humans put into the oceans get concentrated up the food chain — focused and amplified at each level — meaning seafood can carry much higher levels of toxic chemicals than land food, which complicates seafood's reputation as a healthy choice.
“the fish that we eat are themselves carnivores... a focusing and a concentrating effect that all the bad chemicals in the oceans are consumed by the little fishes or the plankton or whatever, and they keep getting consumed by bigger and bigger fish, and they don't go away.”
At the level of quantum field theory, the fundamental ontology is fields not particles; particles are tiny perturbative fluctuations in quantum fields appropriate only when the field is nearly quiet, but fields do non-perturbative collective things (like the Higgs field rolling to its nonzero vacuum value) that are easy to understand from the field perspective and almost impossible if the world were fundamentally particles, which is why particle physicists treat the world as made of fields.
“it is definitely not a theory of particles... if you take the field perspective seriously, there are non-perturbative effects... it's almost impossible to make sense of that if you thought the world were fundamentally made of particles and not of fields.”
Classical systems can be chaotic because their equations of motion are non-linear, allowing small perturbations to grow rapidly (sensitive dependence on initial conditions), but the Schrodinger equation is linear (the wave function appears to the first power on both sides), so the wave function itself never evolves chaotically; chaos must instead emerge in the classical limit of the quantum system, which is why 'quantum chaos' is a non-trivial subject of study.
“In quantum mechanics, the Schrodinger equation itself is linear... So the kind of evolution that is chaotic and classical systems naively doesn't appear at all in quantum mechanical systems. The wave function itself does not evolve chaotically, full stop.”
The Reeh-Schlieder (Taj Mahal) theorem reflects that in the quantum field theory vacuum, the fields in any region of space are entangled with fields everywhere else, and because each region contains an infinite number of field possibilities, a measurement on the fields in one region has a possible outcome that brings into existence literally anything elsewhere — though the probability of such outcomes (like a Taj Mahal on the Moon) is extraordinarily small and not worth worrying about realistically.
“in quantum field theory, in the vacuum state, the fields in any one region of space are entangled with fields everywhere else in any other region of space, all throughout the universe.”
Probability in many-worlds arises from self-locating uncertainty: when the wave function branches, there is a version of you on every branch, and after branching none of them knows which they are on; if those observers try to assign credences in a sensible way, the only reasonable answer is the Born rule (probability equal to amplitude squared), which is the unique set of numbers that add up to one and are conserved over time.
“if you try to answer that question in some sensible way, there is only one possible reasonable answer, which is the Born rule of quantum mechanics, you assign a probability or a credence to be on any one branch that is given by the amplitude of that branch squared.”
Simplicity is a useful guide for dynamical laws (history of science shows looking for simple rules is fruitful) but not for configurations the universe is in; there is no a priori reason to expect the configuration of stuff to be simple, only to expect it to be high-entropy because there are vastly more high-entropy than low-entropy ways to be — so the early universe's uniformity, being low entropy when gravity is considered, demands explanation and is a clue to undiscovered dynamics, not a natural stopping point.
“Simplicity is a very useful thing to keep in mind when thinking about laws of physics... But it's completely different, in my mind, to think about conditions or configurations that the universe can be in.”
Energy is not 'the ability to do work' (since heat in equilibrium can't do useful work); rather, energy is the quantity that is conserved because the laws of physics are invariant under time translation — formalized by Noether's theorem, which associates each symmetry with a conservation law (spatial translation invariance gives momentum, time translation invariance gives energy).
“we call the quantity that is conserved because of that symmetry energy.”
The interference pattern in the double-slit experiment cannot count as evidence for the many-worlds interpretation, because every respectable interpretation of quantum mechanics (Bohmian, etc.) predicts it equally — the Bayesian likelihood of that data is the same across interpretations, so it cannot favor one over another even if many-worlds explains it more attractively.
“the double-slit experiment is predicted by any decent interpretation of quantum mechanics, so therefore it literally cannot possibly count as evidence for many-worlds.”
There is no such thing as 'the speed' you are moving at — relativity says only relative speeds between objects exist. Likewise the expansion of the universe does not have a speed; it has a rate measured in units of one-over-time (roughly the doubling time of the universe), not metres per second, so asking what it is 'relative to' is a category error.
“the expansion of the universe does not have a speed, it just doesn't. It has a rate... that is not a quantity that is measured in metres per second, it does not even have the units of speed, it has units of one over time.”
Because human beings are bounded, finite beings — computationally and informationally circumscribed — we lack the time or capacity for perfect reasoning and will always use heuristics rather than pure Bayesianism or utilitarianism; this finitude is genuinely important and has been put to work in psychology (e.g. the free energy principle as a shortcut to Bayesian updating) and in moral/political philosophy debates about ideal theory.
“the fact that human beings are bounded, finite beings, computationally and informationally circumscribed, is really, really important. We don't either have the time or the capacity to do perfect reasoning about anything, we will always be using heuristics”
Believing a theory is your best current one does not rationally require regarding it as true; rather it requires giving it more credence than any rival you can articulate, while still reserving substantial credence for the set of theories no one has yet articulated — which differs from cases like many-worlds where several competing theories are clearly articulated and can be weighed against each other.
“I give more credence to that theory than to any other theory. However, the total number of other theories might be very large and there might be a large amount of credence I give to some theory and the set of theories no one has articulated yet.”
There is no hard, fast line demarcating scientific from pseudo-scientific theories (the demarcation problem); it is a sliding scale. But Popper was onto two properties good scientific theories tend to have: being definite (incompatible with some outcomes, saying some things happen and some don't, not one-size-fits-all) and being amenable to empirical testing (judged against data rather than intuited from a priori principles).
“I don't think there's a hard fast line... it's more of a sliding scale... there are two properties that good scientific theories tend to have. One is that they're definite, that they're incompatible with something.”
It is impossible to define the size of the universe at the Big Bang itself because that is a singularity where the equations break down, but at any time after the Big Bang the universe could have been infinitely big — this is a feature of general relativity, not inflation, so the whole universe could come into existence infinitely large; statements about the universe being marble-sized at the end of inflation refer only to our observable universe, not the whole of reality.
“at any time after the Big Bang, the universe could have been infinitely big. That's completely compatible with everything we know about general relativity.”
The uncertainty principle cannot explain wave function collapse for several reasons: amplifying a microscopic superposition to a macroscopic one requires no added energy (Schrodinger's cat's awake and asleep states have roughly the same energy); claiming the superposed state 'decays quickly' already presupposes a solution to the measurement problem since superpositions evolve smoothly by the Schrodinger equation; and the uncertainty principle is fundamentally not about measurement at all but about the non-existence of states with simultaneously definite position and momentum.
“to amplify a microscopic quantum superposition to a macroscopic superposition has nothing to do with energy whatsoever. You don't need to put energy into the system at all.”
The phrase 'could have done otherwise' is consistent with a deterministic universe obeying equations of motion, because free will is a useful concept only since human beings are too complicated for us to know their microscopic initial conditions; the right question is whether, among all microscopic configurations compatible with the macroscopic description we have, some would act one way and some another — if yes, that corresponds in higher-level vocabulary to 'you could have acted differently.'
“in the space of all possible microscopic configurations that fit into that macroscopic description, are there some that would act this way and are there some that would act that way? And if the answer is yes, then that corresponds in the higher level vocabulary to, you could have acted differently.”
You cannot ask what a human being is while ignoring emergent properties, because a human being is an emergent property — there are no human beings in the fundamental ontology (quantum field theory, or just a vector in Hilbert space); the question of what a human is is inherently about the relationship of higher emergent levels to lower levels, and you can validly talk about the single-world version of yourself even while believing the fundamental theory predicts copies of you.
“a human being is emergent property. There's no human beings in the fundamental ontology of the world.”
Entropy and weak emergence are related concepts because emergence exploits the compressibility of the microscopic theory — there are many microscopic states that map to the same macroscopic state (e.g. the 10^50 particles in Earth all giving the same center-of-mass motion), and the logarithm of that number of microstates is an entropy; however the relationship is not exact, since some emergent phenomena (Earth orbiting the Sun) do not require increasing entropy, so our general theory of emergence remains primitive.
“The whole idea of emergence absolutely takes advantage of the compressibility of the microscopic theory in some way. And whenever you compress, there's an associated entropy”
Faced with an entity performing physics-defying actions, one should assign low prior to a deity and higher prior to a highly advanced technological civilization, because the deity concept may not even be coherent — it requires violating our understanding of the universe and is built by exaggerating human attributes to infinity (omnipotence, omniscience, omnibenevolence) in ways that are hard to make rigorous, whereas advanced aliens require no such violation.
“I'm not at all sure that the deity concept is even coherent. So I would tend to put low prior probability on the deity idea.”
Dark matter can in principle form black holes but in practice does so far less efficiently than ordinary matter, because dark matter does not couple to electromagnetic radiation, so dark matter particles pass right through each other without dissipating energy and cannot clump; only by riding along with ordinary matter that condenses can dark matter be trapped into black holes in the early universe.
“when dark matter particles come close by, they just go right through each other, they do not dissipate. When atoms bump into each other, they can lose energy by bumping into each other, temporarily getting excited, and then kicking off a photon, by radiating.”
Elaborate unified models like 'dark fluid' should be given much lower credence than the simple standard theory, because the simple theory (dark energy as a cosmological constant, dark matter as cold weakly-interacting massive particles) fits all the data up to small astrophysical anomalies that frequently appear and then go away, so there is no strong killer evidence against it and no need for baroque alternatives until they are well-motivated and empirically successful.
“any of these sort of more elaborate, baroque theories should be given much lower credence than the simple theory, because we have a very simple theory for dark matter and dark energy.”
Any theory of consciousness as a fundamental property of nature must either change the dynamics of the known particles and fields or not; if it does not, the idea has no implications and doesn't matter, and if it does, proponents must show the equations and exactly how the established physics is wrong — and no such serious theory or even a gesture toward one currently exists.
“either you change the dynamics of the particles and fields we know about, or you don't... And if you don't, then I just don't see what it matters... And if you do, show me the equations”
Philosophical zombies are not conceivable if you are a physicalist who holds that consciousness is weakly emergent — just a higher-level way of talking about the collective physical behaviour of matter — because then the same physical collection of matter behaving identically would necessarily be conscious; the zombie argument only works if you assume in advance that consciousness is non-physical, which begs the question.
“zombies are not conceivable. That's what you have to say if you are a physicalist who thinks that consciousness is just weakly emergent.”
The only way to accumulate a billion dollars is for a great many people to send you their money, which only happens if you are embedded in a system with people, roads, education, commerce, and supply chains; therefore it is completely fair for society to ask the rich to pay more taxes in return for maintaining the system that enabled their wealth — no Walmart worker ever saved their way to a billion dollars.
“the only way to accumulate a billion dollars is if a lot of people send you their money. It only happens if you're embedded in a system that allows that to happen”
Modern society has a built-in stratification dynamic that makes it hard to climb out of low income, because life is proportionally much more expensive for poor people than rich people: fixed costs like parking tickets and overdraft fees (literally charging you money for not having enough money) are trivial for the rich but compounding burdens for the poor, so without strong economic and social mobility the system traps people.
“life is much more expensive for poor people than it is for rich people, in a very real sense, because as a fraction of whatever income they have, there are fixed costs that rich people can just laugh off and poor people can't.”
The quantum field is more fundamental than the wave function of a single particle; a single quantum field describes any number of particles including their creation and annihilation, while the single-particle wave function is just a limit you get when restricting to field configurations representing one particle — in that limit the distinction between wave function and field becomes hard to draw.
“the wave function of the field is a much bigger, grander thing. It can describe not just one particle at a time, but any collection of particles, including the creation and annihilation of particles. That's one of the crucial features of quantum field theory.”
Binary black hole mergers serve as 'standard sirens' because they are extremely simple systems with very few degrees of freedom — essentially just two masses (plus spin and tilt) obeying general relativity — so by following the inspiral chirp you can infer the masses, predict the intrinsic gravitational wave amplitude, and combine that with an electromagnetic redshift to determine distance, unlike stars which vary in chemical composition and environment.
“the reason why they're standard is just because there's... All that's happening is gravity with two black holes. There's not a lot of degrees of freedom”
It is more feasible to start from where we are and make things a little better than to define a perfect ideal society and move toward it, because (per Elizabeth Anderson) there may be no such thing as the ideal theory — as society and conditions change, we may learn more and change our minds about what we want society to be — and a similar logic likely applies to everyday personal decisions.
“It's much more feasible to talk about how we can start with where we are and make things a little bit better, than to start with where we are and say, 'Well, over there is the perfect thing, let's move in that direction.'”
The jury system has genuine value despite the fact that jurors are deliberately chosen to be non-experts: it embodies civic participation (reminding citizens that judging guilt is not the domain of distant elites) and, like sortition, people put into deliberative roles usually take them seriously; there is no strong evidence that abolishing juries would yield a better system, since judges and lawyers can be flawed too.
“one aspect that is good, is just the fact that it is participation. Just the fact that it's a reminder to the people on the jury that they play a role in the government”
In the ultra-long-term future, most actual massive particles in the universe will eventually fall into black holes, because most particles in a galaxy are gravitationally bound to it (not on hyperbolic escape orbits) and over enormous timescales will spiral into the central black hole via mixing, gravitational wave emission, and photon cooling — though plenty of photons and some flung-out matter will never fall in.
“most particles in our galaxy are gravitationally bound to the galaxy, they're not just passing through on hyperbolic orbits. And what that means is if you wait long enough, they will spiral into the black hole at the centre.”
In Everettian quantum mechanics, the existence of other branches of the wave function (e.g. one where mass is sent away) does not change the gravitational field you perceive within your branch, because mass exists within each branch and branching does not change the amount of mass in a single branch; the energy you perceive from within a branch is distinct from that branch's contribution to the total energy of the universal wave function, and would otherwise make other branches observable.
“the mass exists within each branch and the branching does not change the amount of mass in a single branch at all.”
In many-worlds, both information and energy are exactly conserved in the wave function of the universe but are not exactly conserved on individual branches where observers find themselves; the information conservation is well known, while the energy variation from branch to branch is less appreciated but true.
“in many-worlds, both information and energy are exactly conserved in the wave function of the universe. They are not exactly conserved on individual branches of the wave function, where observers find themselves.”
Research into consciousness is not an example of strong emergence: by definition zombies behave identically to conscious beings, so every book, discussion, and field of inquiry produced by conscious people could be produced by zombies, meaning the existence of consciousness research does not require top-down causal power; the self-reference problem of modeling yourself within the world (per Jenann Ismael) is a problem of describing the world perfectly given you're in it, not a problem of incompleteness in the laws of physics.
“by definition, zombies behave in precisely the same way as non-zombies do. So every book that a non-zombie would write could also be written by a zombie.”
Most esoteric philosophical questions (which interpretation of QM is right) have little impact on everyday life, but theism versus naturalism does, because meta-ethics (how you justify what is right and wrong) shapes ethics: 'do what God tells you' as a meta-ethical stance yields different behaviour than naturalism — and naturalists have failed to do a good job telling people how to find meaning and live well without God, leaving underlying justifications abandoned without rethinking the resulting modes of life.
“the one example where it does have an impact is in whether God exists, that has a big difference.”
Electrons do not actually have zero size; the claim that they do comes from quantizing the classical theory of a point particle, but if you take quantum mechanics seriously, electrons are wave functions that cannot be localized to a point — their apparent pointlike nature is a feature of the measurement process, not an intrinsic property.
“real world electrons do not have zero size, therefore it shouldn't bother you that they have mass.”
Horizon complementarity — the idea that all information about the apparent external universe is encoded on an observer's cosmological horizon, with no universe extending beyond it — can unify the cosmological multiverse with quantum many-worlds: rather than different physics existing in far-away regions of space, different vacuum states exist as superposed branches of the universal wave function, each describing a finite horizon-sized patch, so the multiverse is instantiated in one patch of spacetime via the many-worlds of quantum mechanics.
“the cosmological multiverse still exists in the presence of horizon complementarity, but it exists because of the many different branches of the wave function”
We could in principle know all the fundamental laws of physics, but we can never acquire total knowledge of science broadly construed, because you can always get more detailed: no human brain can hold the position and momentum of every atom in another human being, so we are forced to coarse-grain into higher-level approximations like psychology and sociology, and there will always be more to learn until we are ourselves as big as the universe — none of us is Laplace's demon.
“A human being cannot carry in their brains enough information to completely describe another human being... that's why we coarse-grain and do higher-order approximations”
It is incorrect to say we move through spacetime at the speed of light; the idea of a speed is distance traveled as a function of time, which simply does not apply to moving through spacetime — and time always passes at one second per second regardless of speed or gravitational field.
“we do not move through space-time at the speed of light... how fast does time go by, it always goes by one second per second. That's the only speed at which time can possibly go by”
Quantum mechanics is not a theory of physics but a framework within which you develop specific theories (the quantum theory of electromagnetism, gravity, etc.), just as classical physics is. There is no experimental problem with quantum mechanics, and the failure to quantize gravity reflects insufficient effort (most attempts still start from a classical theory and quantize it) rather than a flaw in the framework, which has proven extremely robust and resistant to modification.
“Quantum mechanics is not a theory of physics, it's a framework in which you develop theories of physics, just like classical physics is not a theory all by itself.”
When judging an unfamiliar source writing outside their credentialed expertise, you should not dismiss them but should assess their credibility through traits of the person: familiarity with the field, appropriate confidence, clarity about what is known versus speculative, willingness to be wrong (fallibilism), and good-faith versus contrarian/sensationalist motives — rather than relying on a simple algorithm or credentials alone.
“these techniques have more to do with the person talking than the thing they're talking about. Once you know that they're not exactly in their expertise, you shouldn't completely dismiss them, but you should ask questions about how credible they seem.”
History is a set of stories nations choose to tell about themselves from an infinitely large set of possible facts, and nations systematically tell flattering versions that gloss over uncomfortable parts (Japan and Manchuria, France and Napoleon, Britain and colonialism, China and Tiananmen, the US and slavery/Columbus); therefore projects like the 1619 Project that correct a systematic under-emphasis are valuable even if they contain factual errors, because facing uncomfortable history is part of being honest and evidence-based.
“History is a set of stories that we choose to tell, the number of facts about the past that we could potentially write down is infinitely big. The whole thing about history is we pick some of those facts to focus on... and unsurprisingly, countries, nations tend to tell more or less flattering stories about themselves.”
The people who come to mind as 'a physicist' and the issues physicists care about are completely different in the public sphere versus the academic sphere, with only a small overlap; the figures looked up to and questions cared about inside physics departments differ markedly from those discussed on Twitter and in public, so public prominence is a poor guide to influence within the field.
“both the people who come to mind when you say 'a physicist in the modern world', and the issues that come to mind when you say 'What physicists care about in the modern world' are completely different in the public sphere and in the academics sphere. There's some overlap, but it's kind of small”
Because the nucleus is best understood as a swirling mass of quarks without distinct boundaries between protons and neutrons, neutron stars are likely giant blobs of quark soup rather than individual neutrons bumping against each other, since there is a lower-energy wave function for the combined collection of quarks under gravity that the system settles into.
“It does mean the neutron stars are actually giant blobs of quark soup... There's going to be a lowest energy wave function for a certain collection of up and down quarks, that is also feeling gravity”
Graduate school should be challenging but fundamentally fun and rewarding since people enter out of love for the material; prospective students should weigh not only intellectual prestige but how advisors treat students, since abusive advisors exist. However, structural pressures (a tight academic job market, imposter syndrome) are real and unavoidable, and different students need different advising styles, which advisors are never trained to provide.
“this is one of the things you should take very seriously when you're looking at possible graduate schools, not only which is the most intellectual powerhouse, but of the potential advisors of the places you're looking at, do they treat their students well”
Why some ways of solving a problem are easier than others is probably more about features of the problem than features of the human brain; complexity classes concern the minimum number of steps to solve a problem, which is objective and independent of whether a human, dolphin, or computer does it, so we tend to over-credit human psychology and under-credit the fundamental difficulty intrinsic to the problem.
“we tend to not credit enough the fundamental aspects of the problem and we tend to leap too quickly to some particular fact of human psychology”
A PhD is intrinsically worthwhile regardless of whether you ultimately get a professorship — like climbing a mountain but with something useful learned — and the psychological trouble in grad school almost always comes from comparing yourself to other people ('I'm not good enough,' 'I won't get a job'); the antidote is reconnecting with the delight and passion that drew you in and noticing how much more you know than two years ago.
“it's always the comparisons with other people that get you in trouble psychologically here. 'I'm not good enough. I'm not gonna get a job. My advisor is mad at me,' it's always those things that get you in trouble.”
We should develop nuanced views of human beings rather than worshipping or vilifying them; hero-worship extends admiration for one good deed into a general feeling that someone is a good person, which sets you up for disappointment, since a person can both change the world (e.g. Musk converting an industry to electric cars) and be bad in other ways — and the conflation of intellectual disagreement with moral-emotional stances is itself a mistake.
“hero worship involves extending that admiration to a general feeling that they're good people just because they did this one good thing, and that's just setting yourself up for disappointment. A person can both change the world and be bad in other ways”
The stigma against being single is unjustified: society both builds its social functions around couples and assumes everyone should want to be coupled, judging non-conformers as failures — this is a general flaw in how society takes the most common way humans are (like being sexual vs asexual, or wanting children) and treats it as correct while judging all other ways as wrong, so people should be proud of their individual peculiarities rather than worrying about such judgments.
“This is what society does: Society takes the fact that human beings are very diverse, looks at the most common ways that human beings are and judges those to be correct, and judges all the other ways to be wrong.”
There is no fact of the matter about what music is good or bad; what matters about music is what moves you personally, so liking simple bubblegum pop is no worse than liking challenging music like John Cage, and there is no special accomplishment or superiority in liking difficult music — being an aesthetic realist would be strange for someone who isn't even a moral realist.
“I don't think there is a fact of the matter about what music is good and what music is bad. Music is what moves you.”
Caring about truth does not entail a Kantian principle of never lying regardless of consequences; people can fetishize truth-telling too much, and it is fine to tell a kind reassuring untruth (complimenting a haircut already done, telling a young child Santa is real) when the person is seeking reassurance rather than a decision-relevant fact, while being honest when honesty can actually inform a choice.
“there is a danger that people can fetishize always telling the truth a little bit too much... that doesn't mean that there's some sort of Kantian principle that says that you should only ever tell the truth and never lie, no matter what the consequences might be.”
The most important things to know in life are not a set of facts but techniques for learning: gathering information about the world, converting it into models and hypotheses, testing them, and being a good scientist about it — in the broadest sense, including theories of how you should behave, get ahead, and live a meaningful life — because teaching someone to learn lets them acquire knowledge for themselves.
“the most important things to know are how to learn things, techniques for gaining new knowledge, 'cause then rather than just giving you some knowledge... If I teach you to learn things, then you can get knowledge for yourself.”
Twitter is extremely useful as a broadcasting medium for spreading information quickly and linking to longer careful content, but it is poorly suited to in-depth back-and-forth conversations; since following is optional and self-selected, if it consistently drags you down you should reconsider whom you follow or whether to use it at all.
“It's a broadcasting medium... That is the kind of thing Twitter is super duper useful for, linking to things that are longer, more careful things, but not for having in-depth conversations about things.”
Good interviews should not be overly planned; you want a few bullet points to hit, but the specific questions and give-and-take must emerge organically from the conversation, and over-planning (as Carroll did in his early podcasting) produces worse results.
“You want some points you wanna hit, but the specific questions and the specific give and take has to come organically out of the interview.”
Carroll reports being exhausted much of the time despite appearing not to be, attributes his output to working in bursts rather than steady discipline, and says he is not a workaholic and is fairly good at enjoying leisure.
“I'm the kind of person who can be really unproductive for long periods of time, and then a burst of activity and get a lot done.”
Carroll has no sophisticated theory of sandwiches but likes everyday ones (tuna melt, hot dogs, cheesesteaks); he notes a good grilled Chicago hot dog tastes completely different from a lukewarm-water New York street hot dog.
“In Chicago, you grill that hot dog when you order it, and then it's a good hot dog, it tastes completely different.”