What this covers
This is Sean Carroll's October 2021 monthly Ask Me Anything episode, in which he fields questions from Patreon supporters on topics ranging across science, philosophy, and ethics. The format is conversational: Carroll selects and sometimes groups questions he finds interesting, then works through them without a script. Throughout, he applies a consistent intellectual stance—that many celebrated mysteries vanish under scrutiny, and that the real work lies in what remains after metaphysical confusion clears away.
The episode spans a wide territory. On physics, Carroll clarifies how expansion does not affect galaxies, how singularities should be understood as breakdowns in theory rather than features of nature, and why the core theory of the Standard Model plus general relativity fully describes everyday life. On consciousness and identity, he addresses whether the brain receives experience like a radio, whether a Star Trek transporter preserves personal identity, and why coarse-graining resolves apparent paradoxes in neural timing. On fundamental metaphysics, he argues that the Kalam Cosmological Argument rests on folk concepts that fail at cosmic scale, that "why is there something rather than nothing" is not a coherent question, and that many-worlds interpretation requires no spatial distance between branches because space lives within worlds, not between them. On morality, Carroll defends moral constructivism over consequentialism, contending that when utilitarian logic contradicts intuition, the theory is broken rather than intuition wrong. He also touches on how cultural resentment drives anti-scientific sentiment in America, why physicists should interpret credence assignments as confidence rather than certainty, and why synthetic biology rather than traditional nanotechnology will dominate the next technological frontier.
Carroll argues that many grand metaphysical, physical, and ethical questions dissolve when examined carefully: the laws of physics underlying everyday life are completely known, 'why is there something rather than nothing' is not a sensible question, and concepts like causation, the self, and morality must be reframed rather than answered on their own terms.
- The core theory (Standard Model plus general relativity) completely describes the physics of everyday life, even though much complex behavior remains unexplained
- Causation and 'beginning to exist' are informal folk concepts that do not apply to the universe as a whole
- Morality is constructed, not objective, so consequentialism's clean quantification is illusory
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.
Scientists speaking carefully say all life on Earth descends from a last universal common ancestor (LUCA), not that life originated only once; LUCA need not be the first ancestor, so life could have originated multiple times with most lineages dying out or merging, meaning common ancestry does not imply life's origin is extraordinarily rare.
“having one universal common ancestor doesn't mean it was the first ancestor”
It is conceptually wrong to talk about space expanding inside a galaxy; the right description is the spacetime curvature from each star and planet, and an exact solution shows that a spherically symmetric clump of matter surrounded by empty space within an expanding uniform universe experiences literally zero effect from that expansion, not merely a negligible one.
“the galaxy that you made, the clamp of particles is completely unaffected by the fact that the rest of the universe is expanding. A 100% zero unaffected.”
There is no genuinely instantaneous integrated conscious experience; the brain operates at finite, multiple time scales and perceives the world in a coarse-grained way, smearing what we call one moment into a small finite interval, so the space-like separation of neurons poses no real paradox for consciousness.
“it's a mistake to take too literally the idea of instantaneous integrated conscious experience. There is no such a thing.”
The first premise of the Kalam Cosmological Argument, that everything that begins to exist has a cause, fails because causes and effects are informal folk concepts that work in everyday experience but have no guarantee of extending to metaphysical questions about the universe itself; the appropriate question is whether all of reality can be self-contained under simple universal laws of physics, which quantum cosmology suggests is possible.
“This whole first premise of the Kalam Cosmological Argument, everything that begins to exist has a cause. Nope. No, it doesn't.”
Carroll rejects consequentialism because morality is constructed from intuitions, and if a moral theory built on those intuitions reaches conclusions incompatible with them (as utilitarianism does in cases like Peter Singer's views on suffering), the right response is that the theory is bad and needs replacing, not that one must accept the counterintuitive conclusion for logical consistency.
“if the theory leads us to conclusions that are incompatible with the premises on which we built the theory, that just means the theory is bad and we need a better theory.”
The idea that the brain receives consciousness like a radio receiver (Descartes' dualism with a pineal receiver, or Eagleman's analogy) does not work because we know too much about the physics of the particles making up the brain; for such a scheme to operate you would have to change well-established laws of physics governing atoms and electrons, for which there is zero evidence and much evidence against.
“We know too much about the physics of the particles that make up the brain, in order for something like this scheme to work, even if it's encrypted, etcetera, etcetera, you have to change the laws of physics”
Even if effective-field-theory expectations are wrong for some interactions, this cannot give new physics a meaningful effect on consciousness; the only room to change the dynamics of known fields is in higher-order interactions that don't affect everyday life, so making electrons behave differently in a brain than in a rock would require genuinely new fields or a dramatic violation of locality, not just unexpected operator coefficients.
“What they care about is that electrons in the context of a human brain behave differently than the same electron would behave in the context of a rock, and that's not just a matter of the expectations of higher order and non-normalizable operators being off... That's just a total violation of everything we know and love about field theory”
The question 'why is there something rather than nothing' is not a sensible answerable question, like asking for the largest prime number; grammatical form does not guarantee a meaningful answer, and there is simply no such thing as the reason why there is something rather than nothing.
“Just because I can gramatically form a sentence with a question mark at the end doesn't mean it's a sensible question to ask about the universe”
Concluding that God does not exist is only the beginning, not the end, of the intellectual work; naturalists have spent enormous effort arguing against God but have not given adequate answers to what follows, such as the meaning of life, free will, and how to be moral, which is why moving naturalism forward is the real project.
“Once you say God doesn't exist, you're not finished.”
Using a Star Trek transporter is not dying because there is no nonphysical essence attached to the body; the world is just the stuff in it and what that stuff is doing, so the reconstructed person continues your personal identity for the same reasons your future self does in ordinary continuous existence.
“I don't think there are any such essences, I think that the world is the stuff in it and what the stuff is doing.”
Epistemic structural realism holds that across scientific revolutions we throw away the ontology (the fundamental stuff) but preserve and improve the structures of what relates to what, including the mapping between formalism and observation, so phenomena like the Earth orbiting the Sun survive even as classical mechanics gives way to relativity or quantum mechanics; the ontic version, claiming structures themselves exist without things, Carroll finds unintelligible.
“the structural realists say that the thing that is real are the things that are preserved through these transitions from one theory to another, and they talk about in terms of the structure of what relates to what.”
A good Bayesian assigns near-zero prior to almost nothing logically coherent, and stating a confident answer (like 'dark matter exists' or 'many worlds is correct') reflects high credence, not certainty; it is a mistake to attribute certainty to a speaker simply because they make a claim without footnoting their numerical credence.
“it will be very tedious indeed to go through life mentioning your numerical Bayesian credence to every claim that you make, even if that credence is 99%.”
Once you communicate publicly, you will constantly be misrepresented, often as holding the opposite of your actual views, because everyone has finite capacity and fits others' ideas into pre-existing mental categories that do violence to the original; the only solution is for everyone to make good-faith efforts at accuracy, and to pick your battles rather than chase every misrepresentation.
“once you do some communication in the public sphere, once other people read what you write or listen to what you say and then talk about it, you will constantly be misrepresented.”
Physics is remarkably unified because it shares an underlying core (the Standard Model plus general relativity) and common methodologies, so a result violating that core is either wrong or Nobel-worthy rather than just 'a different field'; this unification stems from physics being simpler than other sciences, which lack that privilege.
“physics is remarkably unified considering how much ground it covers, and that comes down to the fact that physics is pretty simple compared to other things”
Much anti-scientific sentiment in the US is driven not by evaluation of evidence but by cultural resentment, with subpopulations reacting against being lectured to by elites; thus arguing creationism or flat-earth on the scientific merits is counterproductive, and the better approach is to ask what makes people distrust the scientific story.
“the desire to get creationism taught in schools was not really about creationism, it wasn't even really about religion, it was about asserting the right to determine for themselves what they wanted taught in their local schools, rather than being told by some government bureaucracy or some academic elites”
In physics a singularity is best understood as a place where the theory breaks down rather than a real feature of nature, so the meaningful question is not 'what happens at a singularity' but 'what happens instead of a singularity' where the theory ceases to be accurate.
“usually the way that professional physicist thinks about a singularity is as a place where your theory is breaking down.”
The curvature of space is completely independent from the curvature of spacetime: a universe with a positive cosmological constant approaching de Sitter space can simultaneously have spatial slices that are positively curved, negatively curved, or flat, and in pure de Sitter space the same spacetime can be sliced to yield either flat or positively curved spatial slices.
“The curvature of space is completely independent from the curvature of space time.”
Information in the physical sense is a correlation between physical configurations requiring no interpreter, so an untranslatable manuscript that says something correct about the world contains information regardless of whether anyone can read it; a separate notion of accessible or usable information can be far smaller, which is why burning a book destroys information in practice though it is conserved in principle.
“one definition of information is a correlation between one physical configuration and another one.”
When a book is burned, the universe as a whole retains the complete microscopic information to reconstruct it, but this information is inaccessible to us because measuring quantum systems alters them; macroscopic irreversibility reflects this inaccessibility, and because a high-entropy macrostate contains far more microstates, a typical microstate evolved backward would not reconstruct the book but would tend toward higher past entropy.
“What we mean when we say the information is still there, is that in the universe as a whole, the universe knows the complete state of the universe, even though it's inaccessible to us because of the uncertainty principle”
Quantum fields do not generally have analytic properties; analyticity may hold in special states like the vacuum of a quantum field theory but not in an arbitrary state, so once you have a black hole with information dumped into it there is no reason a function's behavior in one region tells you its behavior elsewhere, defeating the proposed analytic-continuation resolution of the black hole information puzzle.
“I don't think the quantum fields have analytic properties, there's no reason for them to. It may be that in some particular state, quantum fields have analytic properties, but in an arbitrary state, they don't.”
In the many-worlds interpretation there is no physical distance between branches, because space lives within each world separately rather than worlds living within space; the wave function of the universe is fundamental and not located in space at all, so each world has its own copy of space with no spatial separation between them.
“there is literally no such concept as the distance between the worlds. It's the other way around, space lives within each world separately, it's not that the worlds live within space.”
Carroll's self-locating-uncertainty approach to many-worlds probability gives the same Born-rule answer as the Deutsch-Wallace decision-theory approach but is more satisfying because it pinpoints why there is probability at all, namely that after branching you don't know which branch you are on, whereas the decision-theory approach only shows that if you reason probabilistically it must be the Born rule.
“Chip and I following on ideas from Lev Vaidman pinpoint where the necessity a probability is, namely in the self-locating uncertainty that you don't know which branch you're on.”
Tiny technologies will be hugely important and underestimated in the near-to-middle-term future, but the biggest impacts will come from synthetic biology rather than traditional metal-machine nanotechnology, because at the nanoscale the appropriate way to function is to look like a cell rather than a miniaturized robot, as exemplified by repurposing CRISPR from bacteria.
“the biggest impacts are not gonna come from what we traditionally think of as nanotechnology, but from what we traditionally think of as synthetic biology”
In most areas of science working physicists routinely discover new facts about the universe, but in fundamental or elementary physics the basic laws have been in place since the 1970s, so any genuinely new fact about the fundamental laws would now win a Nobel Prize, with recent measurements (Higgs mass, cosmological constant, neutrino masses) being confirmations of ideas already proposed decades earlier.
“in fundamental physics, if you discovered a new facts about the universe you would absolutely win a Nobel Prize.”
It is harder but not impossible to enter professional physics through self-study, because you can demonstrate proficiency by a killer physics GRE score to enter grad school, or by publishing good papers in journals like Physical Review to be hired without a PhD; the obstacle is that formal training also imparts informal folk wisdom from environment, advisors, and collaboration.
“if you can convince them that you would do well in grad school, then they would often be happy to take a chance. Remember, you can apply to 50 grad schools and you only need one to take you.”
Individual choices like not flying have strictly zero effect on the climate because the same planes fly regardless; solving the climate crisis requires collective action through governments and international cooperation, setting up systems (like making fossil fuels expensive and incentivizing alternatives) so that the choices making people happiest are the ones good for the planet, rather than hectoring individuals into good behavior.
“me not flying would have strictly zero effect on the climate... 'cause the same planes would fly, like my choice not to fly, would not affect the flight of the planes.”
We cannot know for certain whether any current theory is the fundamental description of reality, and right now we definitively lack a theory of everything because general relativity and quantum mechanics are incompatible with each other in regions where both should hold; the goal is not necessarily to find a deepest level but to keep finding deeper levels or a covering theory.
“we do not have that underlying theory of everything because we have general relativity and we have quantum mechanics. They're incompatible with each other.”
Gravitational waves were expected to and do propagate at the speed of light because they are massless and obey a wave equation with the same differential structure as electromagnetic waves; the special speed is a built-in feature of spacetime that massless phenomena travel at, not something special about light itself, and giving gravity or light a different speed is very difficult given what we know.
“there is a special speed built into the fabric of space time, the speed at which you can't ever travel the limiting speed, if you are a massive particle or object”
A field exists not because it currently has a nonzero value but because it could in principle change and interact with other fields; a field whose value is necessarily zero everywhere might as well not exist, while one that merely happens to be zero in a region but could be excited still exists.
“The whole point of fields is that they can in principle change and interact with other fields.”
You could never see gravitational waves rippling off colliding black holes or neutron stars because seeing means detecting electromagnetic waves and gravitational waves emit zero light; you could only observe surrounding objects moving relative to each other, but you would be seeing those objects, not the gravitational distortions directly.
“It's gravitational waves, the only thing you can ever see are electromagnetic waves, that's what it means to see”
Branches of the wave function become genuinely distinct worlds at the branching (decoherence) event rather than always having existed and merely differentiating over time, because poking one purported world before branching would be noticeable in the other parts of the wave function, whereas after branching changing one world has no observable effect on the others.
“if you went in and poked one of those purported worlds before the branching event happened before decoherence happened, that would be noticeable in the other parts of the wave function”
Whether the number of many-worlds branches is finite or infinite is unknown because the assumption that an electron is a point particle in a continuum (which would give infinite worlds) is only a low-energy approximation; in quantum gravity, where Hilbert space may be finite-dimensional and measurements break down at quantum-gravity scales, the continuum picture fails, so the answer depends on physics we don't yet have.
“you've assumed a model of electrons, namely, you've assumed that the electron can be classically described as a point particle moving in a continuum, and that might not be true, indeed in quantum gravity it wouldn't be true.”
In many-worlds the amplitude has a real-world effect: it gives the likelihood of which branch you find yourself on, because after branching the people on different branches are different people, so each one can ask which branch they are on and assign a probability equal to the amplitude squared.
“if I exist after the branching and I know that I'm on one branch but not the others, which one am I on? And I assign a probability, and that probability is given by the amplitude squared.”
Public theism-versus-atheism debates serve a useful purpose not of finding truth or persuading the opponent, who is set in their ways, but of educating undecided members of the audience about ideas the speaker already holds.
“there is a useful purpose to these debates, but it's not finding the truth. It is educating the audience about ideas you already have.”
Standard quantum mechanics textbooks fall short by treating the Schrodinger equation for single-particle systems repeatedly while mentioning entanglement only in passing, which does a disservice to what makes quantum mechanics distinctive; a better textbook would foreground entanglement, decoherence, and density matrices, partly because these underpin modern applications like quantum computing.
“very often, quantum textbooks are just excuses to solve the Schroedinger equation over and over and over again, most often for single-particle systems, and the notion that particles can be entangled is mentioned and then slid by. That's terrible.”
Competing philosophies are often hopefully assumed to be one at heart, but generally they are genuinely different, just as consequentialism, deontology, and virtue ethics differ; Confucianism emphasizes social roles and hierarchy while Daoism emphasizes individual self-cultivation, and although they can be compatible they are not simply the same path.
“there is a sort of hopeful tendency... to say like, aren't they all one at heart? And I think that the general answer... is no”
The fact that the early universe had a low entropy is a really important problem to understand that Carroll's cosmology colleagues do not always treat as significant, exemplifying a problem he considers of utmost importance that the consensus underrates.
“I still think that the fact that the early universe had a low entropy is really, really, really important to understand, and my cosmology buddies don't always go along with that”
Whether quantum branching produces radically diverse histories depends crucially on how far back the branching occurred: ancient branchings (e.g., inflationary fluctuations seeding which galaxies form, or genetic recombination) can yield wildly different presents, but recent branchings affecting only the actions of 1930s-1940s leadership leave much smaller room for outcomes like the Nazis winning World War II.
“as you think about branchings of the wave function that go earlier and earlier in time, then you can have a tiny fluctuation, a tiny difference at a very early time lead to an enormous difference in the present era.”
There is a puzzle that entropy increases while entropy also measures our ignorance and we continually learn about the world (so a good Bayesian seemingly sees a reversed arrow of time as entropy falls); Carroll's Bayesian Second Law resolves this by deriving a quantity, involving relative entropy, that increases even when measurements that reduce uncertainty are included.
“how do you deal with the fact that on the one hand, you say entropy increases, on the second hand, you say that entropy is a measure of our ignorance, and on the third hand, you learn things as you look at the world”
Using the word 'energy' in a psychological sense (extroverts gaining energy, introverts being drained) is perfectly acceptable even though it corresponds to no physical energy flowing into the body, because the metaphorical purpose is clearly served and everyone understands the meaning.
“it doesn't correspond to anything to any physical stuff going into their bodies or anything like that, it has to do instead with intricate questions about their physiological state”
Self-learning professional physics is more doable than ever thanks to online courses and lecture notes, requiring two things: a curriculum (algebra through calculus, differential equations, linear algebra, then classical mechanics, electromagnetism, quantum mechanics, statistical mechanics and thermodynamics) and the motivation to sustain it, with the main difficulty being that learning retrains the brain and requires breaking it into bite-sized chunks.
“really what you need... You need two things, You need a curriculum, so you need a list of subjects to learn, and then you need the actual motivation to do it.”
Saying the laws of physics underlying everyday life are completely known is fully compatible with needing more physics, because the completeness claim applies only to the core theory of quantum field theory governing everyday phenomena, not to how those fields combine to make complex things like books and people, which remains far from established.
“the kind of physics that we need is not of the form laws of physics underlying everyday life at the level of the core theory of quantum field theory, that part of physics is done.”
The swampland-derived argument that string theory is incompatible with a positive cosmological constant is not airtight; the game of analyzing string-theory compactifications is hard enough that sweeping claims about what is and isn't possible are premature, so Carroll's credence has not shifted much, and dynamical dark energy like quintessence remains a less-favored alternative since it would likely already have been detected.
“the game being played of looking at compactifications of string theory is a sufficiently hard game to play that making these very, very general statements about what is and is not possible, is a bit premature.”
In Carroll's framework Hilbert space is the starting point and spacetime emerges from it, so gravity does not act on Hilbert space; rather, only finite-dimensional Hilbert spaces can support general-relativistic gravity, while infinite-dimensional ones either lack gravity or have it only in forms where local spatial regions are described by finite-dimensional pieces.
“the Hilbert space is always more fundamental than spacetime or anything like that, at least in my way of thinking about it”
There are no real tricks to make learning challenging concepts feel less like a chore; learning hard subjects requires retraining your brain along unnatural paths through practice, thought, repetition, and working problems, though breaking it into bite-sized sub-subjects and treating each as an accomplishment can provide a sense of progress.
“when you're learning something like this, You're retraining your brain, your brain doesn't naturally go down those roots that you need to go to learn about modern physics”
We are more likely to blame individuals and less likely to blame systems than we should be because it is emotionally easier to personalize blame, but in complex systems individual impact is limited; nonetheless powerful actors do have more impact and it is legitimate to assign them the blame they deserve.
“too often we personalize the blame because it's much easier emotionally to place blame on a person than it is on institutions and systems.”
Whatever humanity's track record on good or bad intentions, our track record for doing whatever we become technologically capable of doing is nearly 100%, so we should expect that interventions like seeding lifeless planets, designer babies, or uploading consciousness will eventually happen.
“their track record for doing what they're capable of doing is almost 10% we do it. We dropped atomic bombs on cities full of people. That's how human beings roll when we derive some new technological capability, no matter how horrific the consequence is, we do it.”
Much of the modern weirdness around marriage stems from conflating two very different uses, the romantic commitment and the legal arrangement, and these could be productively separated so marriage could be purely romantic while legal protections (like medical decision rights for a partner) are handled through other mechanisms.
“a lot of the weirdness about marriage in the modern world is because these are two very different uses, the romantic use and the legal use, and the legal use is not to be taken lightly.”
Divorce is made unnecessarily difficult because society treats it as something bad that ought to involve suffering, but the mechanism to end a marriage should be simpler and one person wanting out should generally suffice.
“we have some feeling that divorce is somehow bad and therefore, if you're going through it, you should suffer, it should be difficult to go through it. And I'm against that.”
In the many-worlds interpretation, copies of a person on different branches share a common past but not a common future, and since they cannot affect one another, each should be regarded as a separate self; the everyday operation of personal identity is unchanged because we never observe the other branches.
“they share a universal past a common past, but they don't share a common future, the things that will happen to them in the future are distinct. So that changed the fact that they do not share a common future, to me means they're separate people”
American exceptionalism in anti-science resentment may stem from the US being unusually heterogeneous, a large mostly non-native population with a complicated regional history, so subpopulations do not all identify with the rest of the country and resist being told what to believe.
“the people who are in the United States today are the overall majority of them, not descended from families who were here in the United States 500 years ago”
A wormhole-based teleportation that connects two locations through spacetime would be more efficient than a Star Trek-style destroy-and-recreate transporter, because the wormhole lets the existing laws of physics carry the object's atomic configuration along for free, whereas a transporter must move each particle by hand.
“if you go to the effort of building a wormhole or a bridge through spacetime, then the sort of nuanced effort of moving each individual atom and molecule in the body of the cow is done by the laws of physics, it goes along for the ride.”
Carroll assigns low credence to there being horrible inner experiences under successful general anesthesia, reasoning by analogy to dreamless sleep that anesthesia likely puts you so deeply under that you have no experiences at all, and that whatever experiences might occur would resemble dreams more than tortures, so this should not deter beneficial surgery.
“the general anesthesia puts you under so deeply that you don't have experiences, you don't have anything that you would recognize as an experience, just as you don't when you're asleep and not dreaming.”
If a planet is genuinely devoid of life, putting the raw materials of life onto it harms no one, since there is literally no party who could be harmed by such an intervention.
“if it's literally another planet that doesn't already have life on it, I don't see the harm, I don't see literally who is being harmed by us doing some dramatic intervention to it.”
Carroll's writing process emphasizes extensive thinking, outlining, and conceptualizing before writing anything, so that drafting requires little revision; he writes start-to-finish with a clear outline and title in mind, contrasting with the common free-writing approach of producing rough drafts and then heavily cutting.
“I really try to think in the most organized way possible about what I wanna say, where I'm going, and then once it comes to actually writing it... I don't need to revise very much”
The single best thing about the Boston area is that it is the best university town around, with an amazing concentration of smart people devoting their lives to intellectual pursuits across Harvard, MIT, BU, BC, Tufts, Northeastern, and Brandeis.
“The amount of smart people devoting their lives to intellectual pursuits is just amazing here in the Boston area, that's my favorite thing”