What this covers

This February 2023 Ask Me Anything episode features Sean Carroll fielding questions from Patreon supporters on topics spanning physics, philosophy, technology, and ethics. The episode touches on fundamental questions about the nature of the universe—whether we live in a black hole, how spacetime emerges from quantum mechanics, and what many-worlds implies for probability and ethics—alongside more immediate puzzles like dark matter's role in galaxy formation and whether current AI systems possess genuine intelligence. Carroll's responses cluster around a few recurring themes: that apparent cosmic mysteries often reflect mismatched human expectations rather than genuine strangeness, that real reasoning operates under Bounded Rationality and Satisficing rather than perfect optimization, and that higher-level descriptions like spacetime and branching worlds are Emergence from deeper quantum structure.

The episode ranges widely across physics detail. Carroll clarifies misconceptions about black holes and mass generation, explains why gravity cannot remain classical while matter is quantum, and challenges the coherence of downward causation with quantum field theory. He addresses Naturalness in cosmology—discussing fine-tuning, dark energy, and inflation—arguing that when the universe seems unnatural, the fault lies with our expectations, not reality. On artificial intelligence, he argues that large language models like ChatGPT produce fluent fabrication because they lack genuine concepts and world models, only pattern prediction. The discussion of many-worlds cuts deeper, treating Decoherence as the mechanism that converts branches from potential to actual, and exploring how probability arises from self-locating uncertainty in a deterministic wave function. Carroll also engages with applied ethics—defending empathy against critics, examining whether utilitarianism and deontology are truly distinct, and testing how moral reasoning holds up under many-worlds considerations.

Sharpest takeaway

Carroll argues that current large language models only mimic intelligence by predicting word patterns without any conceptual understanding, and more broadly that many apparent mysteries in physics (low entropy, fine-tuning, spacetime emergence, many worlds) reflect mismatched expectations rather than the universe being strange, while emphasizing that real reasoning is bounded and approximate.

  • ChatGPT produces fluent but factually fabricated answers because it lacks concepts and a world model
  • Physical 'preposterousness' (fine-tuning, low entropy) signals that our expectations are wrong, not the universe
  • Spacetime, the self, and worlds in many-worlds are emergent, approximate descriptions of an underlying quantum wave function

The claims · ranked52 claims · weighted by value

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.

0.86

We do not live inside a black hole: our universe shows no future singularity and no outside region to escape to, the two defining black hole properties. The singularity theorems are time-symmetric, but our universe had its singularity (the Big Bang) in the past, making it conceptually closer to a white hole than a black hole.

factualhigh valueestablishednovelty 3/4durability 4/4· Sean Carroll

“If anything, our universe is like a white hole. A white hole is the time reverse of a black hole and it has a singularity in the past from which you can escape.”

0.86

Downward causation—the idea that higher-level context (being in a rock vs a brain) affects how fundamental particles behave—is entirely incompatible with quantum field theory, because field equations of motion depend only on the values and derivatives of fields at exactly the same point in space, so nothing far away can affect a field's evolution at a point.

factualhigh valueestablishednovelty 3/4durability 4/4· Sean Carroll

“It is very clear, completely unambiguous, that nothing that is going on at other points in space far away can possibly affect the equation of motion for the field at this point.”

0.86

Although physics often treats one part of a system classically and another quantum-mechanically, gravity cannot fundamentally remain classical while matter is quantum, because it is a categorical type mismatch (like adding a number to a vector): general relativity needs a definite energy at each spacetime point to determine curvature, but quantum mechanics has no pre-existing energy value, only measurement-dependent outcomes.

factualhigh valueestablishednovelty 3/4durability 4/4· Sean Carroll

“It is literally just not possible for it to work that way... It's like saying, 'Can you add a real number to a vector?' You're just trying to combine two different types of things.”

0.81

Tiny quantum fluctuations can grow into macroscopic differences across wave-function branches when classical chaotic behavior is present, as with Saturn's chaotically tumbling moon Hyperion (which Zurek showed would smear into a wave-function blob if not constantly decohered by its environment), and through mutations that alter evolutionary history, though whether this matters for one human life over five years is an unanswered quantitative question.

causalhigh valueestablishednovelty 3/4durability 3/4· Sean Carroll

“when there is classical chaotic behavior, then guess what? A tiny variation is really important. Even a variation that is small enough just to be due to quantum effects.”

0.81

There is no rigorous division of entropy into gravitational and non-gravitational forms; the correct statement is simply that the early universe had low entropy, and the intuition that the smooth, thermal early universe was low entropy comes from the fact that with gravity, densely packed matter is high entropy when lumpy (black holes) and low entropy when smooth.

factualhigh valueestablishednovelty 3/4durability 3/4· Sean Carroll

“Let me emphasize that there is no clear rigorous division of entropy into gravitational and other forms... The point is that, the correct absolute thing to say about the early universe is that the entropy was low, full stop.”

0.80

Even if the correct action is expressible as optimizing a number, humans never actually do this perfect optimization because we are bounded in both collecting and processing information; in practice we always satisfice (settle for good enough), as reflected in approaches like Friston's free energy principle that approximate Bayesian updating quickly.

factualhigh valueestablishednovelty 2/4durability 4/4· Sean Carroll

“We're bounded in our abilities to both collect information about the world and then to process that information. So in practice, even if we think that the right thing to do is perfectly well expressed by optimizing some number, in practice, we never really do that.”

0.79

Dark matter affects galaxies but not planetary orbits because dark matter particles lack electric charge: ordinary particles can emit massless photons of arbitrarily small energy, dissipating energy and sticking together to form stars and planets, whereas dark matter cannot radiate energy away, so it passes through itself and remains diffuse, leaving negligible density in the solar system.

causalhigh valueestablishednovelty 3/4durability 4/4· Sean Carroll

“when different particles come together, they can dissipate, they can hit each other and then lose energy by emitting some photons. And that makes them stick together. Two dark matter particles, which as far as we know don't couple to anything like electric charge... just pass right through each other.”

0.79

It is false that the Higgs field is responsible for mass in general; mass is simply rest energy divided by c-squared, and most of a proton or neutron's mass comes from quantum chromodynamics (the strong force), not the Higgs, which only gives mass to certain elementary particles like quarks, leptons, and W/Z bosons whose symmetries would otherwise keep them massless.

factualhigh valueestablishednovelty 3/4durability 4/4· Sean Carroll

“Protons and neutrons, which are the particles that make up most of your mass as a human being, get most of their mass from quantum chromodynamics, from the strong nuclear force. Nothing to do with the Higgs field at all.”

0.79

The puzzle is not that we expected black hole information to scale with volume; rather, for an ordinary thermodynamic system at maximum entropy (where gravity is unimportant), entropy is proportional to volume, but a black hole in equilibrium has entropy proportional to its area instead, and we don't know why this differs so dramatically from the volume scaling.

factualhigh valueestablishednovelty 3/4durability 4/4· Sean Carroll

“when gravity is not important, entropy is proportional to volume in equilibrium. Black holes are in equilibrium, their entropy is not proportional to their volume. What is going on? That's really the question being asked.”

0.79

Although gravity is geometry classically, quantizing it yields a graviton because in the weak-field limit curved spacetime is approximated as flat spacetime plus a gravitational field, and quantizing the small perturbations of the metric tensor inevitably produces particles, just as quantizing any field produces particles via quantization conditions on vibration energies.

factualhigh valueestablishednovelty 3/4durability 4/4· Sean Carroll

“the small perturbations of the metric tensor that describe gravity in the weak field limit are something we can very easily quantize. And out pops gravitons, out pops particles. It's almost inevitable.”

0.78

The distinction between optimizing utilitarians and rule-following deontologists or virtue ethicists is partly fake, because any decision procedure can be formally mapped onto maximizing some quantity (e.g., assigning infinity to true statements and minus infinity to lies), so the apparent contrast between optimizing a number and following rules is one of convenience, not substance.

factualhigh valuecontestednovelty 3/4durability 4/4· Sean Carroll

“anything that you do, any procedure that you can invent for making decisions can be thought of as maximizing some quantity.”

0.78

When a feature of the universe (like dark energy's coincidence problem, the cosmological constant problem, the hierarchy problem, or inflation) seems unnatural or preposterous, it is never a criticism of the universe but a reminder that our expectations—set by some subset of what we observe—are wrong, signaling something we don't yet understand.

normativehigh valuecontestednovelty 3/4durability 4/4· Sean Carroll

“whenever there's something that seems unnatural to us, we are making a mistake. It's not a criticism of the universe, when I say the preposterous universe, it's a reminder that there is something that we don't yet understand.”

0.75

Real progress, including in AI research, is non-linear—poking along slowly for a while and then leaping forward—even though in historical hindsight we tend to perceive progress as having a steady rate, so current AI advances are genuinely important and not yet finished even if they aren't true intelligence.

factualhigh valueestablishednovelty 2/4durability 3/4· Sean Carroll

“we tend to see progress as something that has a rate when we look back on it historically from a remove, but real progress is very non-linear”

0.75

We neither assume nor have evidence that the accelerating expansion characterizes the whole universe; within the observable universe acceleration is a data-based conclusion since things look the same everywhere, but beyond it we have no evidence and need not assume uniformity, as eternal inflation scenarios suggest radically different conditions outside.

factualhigh valueestablishednovelty 2/4durability 3/4· Sean Carroll

“Within the observable universe, things seem to be the same everywhere... including the fact that the universe is accelerating. So that's a databased conclusion. Beyond the observable universe, we neither have evidence that it is accelerating in the same way, nor do we assume”

0.74

A black hole is a region of empty spacetime, and its 'spin' means the metric tensor field defining its geometry carries angular momentum (inherited from the collapsing star); this makes the hole oblate (the Kerr metric), tilts the local light cones so infalling objects are dragged around with the spin, and extremal black holes 'spin at the speed of light' but cannot physically exist.

definitionhigh valueestablishednovelty 3/4durability 3/4· Sean Carroll

“the thing that has angular momentum is really the value of the metric tensor field.”

0.74

Inflationary models are not strongly contingent on supersymmetry; while the inflaton potential must be fine-tuned (shallow slope, ~60 e-folds), supersymmetry can be broken at a high scale and still help, so the failure to find supersymmetry at the LHC affects the hierarchy problem (which needs SUSY broken near the electroweak scale) but not inflation.

factualhigh valueestablishednovelty 3/4durability 3/4· Sean Carroll

“I don't actually think it's true that inflationary models are very contingent on Supersymmetry.”

0.73

Contrary to Paul Bloom's Against Empathy, more empathy is crucially important because people too easily use rationality badly—thinking they are being rational while ignoring important aspects of a problem—and empathy helps ensure we consider all relevant perspectives, especially of people in very different circumstances.

normativehigh valuecontestednovelty 3/4durability 3/4· Sean Carroll

“It is a rather... It's not that rationality is hard, it's that it's too easy to use rationality badly. And I think that empathy can be something that helps us get over that”

0.73

In many-worlds, before measurement there is nothing to be probabilistic about; probability is the self-locating uncertainty of future copies of you who, after decoherence, don't know which branch they are on, and even though everything is deterministic at the wave-function level, it is rational and correct to act and speak as if there are objective probabilities.

definitionhigh valuecontestednovelty 3/4durability 3/4· Sean Carroll

“What the probabilities are, are the probabilities of which branch we're on in the circumstances when we don't know. The self-locating uncertainty”

0.73

For most realistic moral theories, distributing money across many-worlds branches is equivalent to a single world with probabilistic outcomes; giving 98 people a million, one nothing, and one twenty million is the same analysis as a 98% chance of a million, 1% of nothing, and 1% of twenty million, so many worlds is a distraction in such ethical analysis.

factualhigh valuecontestednovelty 3/4durability 3/4· Sean Carroll

“I don't think that the many worlds thing matters that much in this kind of analysis. I think many worlds is just a distraction. Just think of it in terms of probabilities, and I think it's exactly the same analysis”

0.73

The Copernican Principle that we are not special is essentially nonsense because it is utterly ill-defined whether 'equal' should apply per year of cosmic history or per person/intelligence, and whether a collective intelligence counts once or many times, so we should not decide in advance what the distribution should be and call reality remarkable.

normativehigh valuecontestednovelty 3/4durability 3/4· Sean Carroll

“I think the Copernican Principle is pretty much nonsense, to be honest”

0.73

Spacetime is not emergent from any familiar substance or stuff; the fundamental description of nature is a quantum wave function evolving via the Schrodinger equation, and spacetime obeying the Einstein equations is a higher-level coarse-grained description of that wave function, analogous to fluid mechanics emerging from atomic motion.

factualhigh valuecontestednovelty 3/4durability 3/4· Sean Carroll

“the fundamental description of nature is a quantum state, a wave function, evolving in time according to the Schrodinger equation. And there is some higher level description of that, some coarse grained description that looks like spacetime obeying the Einstein equations.”

0.73

Branches in many-worlds should be thought of as one world that branches into many when decoherence occurs, not as pre-existing identical worlds, because before decoherence interference is possible; worlds are convenient higher-level human constructions that become obvious when macroscopic branching happens, and ambiguous microscopic cases simply don't matter to the fundamental ontology.

definitionhigh valuecontestednovelty 3/4durability 3/4· Sean Carroll

“I've always been of the opinion that worlds are convenient higher level human constructions. That are very convenient, but they're very obvious when they happen, when the branching happens.”

0.73

Ocean and land life don't count as separate samples for estimating how often civilizations arise, because the two biospheres share a last universal common ancestor and are deeply intertwined; moreover, one successful civilization or life-form can crowd out all others by occupying ecological niches first, so we still have a sample size of one civilization per planet.

factualhigh valuecontestednovelty 3/4durability 3/4· Sean Carroll

“one successful development of civilization could easily crowd out all other possibilities... there are ecological niches and whoever starts first is likely to win.”

0.73

There is a parallel between many worlds and Carroll's approach to the past hypothesis: in both, an apparent enigma (wave function collapse, or the low-entropy beginning) dissolves when our universe is recognized as part of a larger system—decohered branches all exist, and baby universes spawning from a larger system explain low entropy—reflecting that mysteries arise from too-narrow expectations.

factualhigh valuecontestednovelty 3/4durability 3/4· Sean Carroll

“putting the universe that we observe into a larger context can be one way of doing that.”

0.73

The self is real but not fundamental—it appears nowhere in the laws of physics yet is useful for describing the human-scale world, so it is emergent; the persistence of the self through time was always only an approximate notion, since you are never exactly the same person you were a minute or year ago, though there is good continuity.

factualhigh valuecontestednovelty 3/4durability 3/4· Sean Carroll

“the idea of a self is something that is nowhere to be found in the fundamental laws of physics, but it's clearly very useful as part of our description of the everyday human-sized world, so therefore, it is emergent.”

0.73

The cross product of two 3D vectors is defined as a third vector perpendicular to their plane whose magnitude equals the product of the magnitudes times the sine of the angle between them, which equals the area of the parallelogram the two vectors form; it is perpendicular because we define it that way, and there are only a few coordinate-independent ways to construct a vector from two vectors.

definitionhigh valueestablishednovelty 2/4durability 4/4· Sean Carroll

“the magnitude of the cross product vector is just the area of a little parallelogram that the two vectors describe. Why is it always perpendicular? Well, we defined it that way.”

0.73

It is wrong to think gravity makes things fall to the ground on Earth but orbit in space; gravity is identical in both cases. Objects hit the Earth because the Earth is a large solid body and ordinary velocities are far below escape velocity, but if Earth were compressed to a black hole of the same mass, a baseball could easily orbit because the body would be small enough not to be hit.

factualhigh valueestablishednovelty 2/4durability 4/4· Sean Carroll

“Gravity is exactly the same in both conditions. What's different in both conditions is the set of possibilities that you are familiar with”

0.73

Traveling at high speed relative to the cosmic microwave background makes its temperature appear anisotropic—hotter ahead, colder behind—and this actually occurs: there is a dipole temperature anisotropy in the CMB (about 10^-3, versus 10^-5 intrinsic variations) attributed to our motion at roughly 300 km/s relative to the unique CMB rest frame.

factualhigh valueestablishednovelty 2/4durability 4/4· Sean Carroll

“There is a dipole temperature anisotropy in the cosmic microwave background that is big... And we attribute that dipole to our motion with respect to the cosmic microwave background rest frame.”

0.68

The US founders deliberately separated powers so divided government would force compromise and reflect the people's incomplete preference for either side, but in the present situation game-theoretic incentives mean politicians stay in office by opposing everything the other party does rather than compromising, so a parliamentary system would trade these problems for new ones and is politically unattainable anyway.

factualhigh valuecontestednovelty 2/4durability 3/4· Sean Carroll

“the original idea of the US Founders was that divided government could be good. It could actually reflect the people's inability to choose what to do.”

0.66

Burning a book differs from throwing it into a black hole because in principle we can use the Schrodinger equation to predict exactly what radiation and ash a fire produces, preserving the information, whereas Hawking's calculation gives only thermal radiation with no apparent relationship to the book's information beyond its mass, which is the essence of the information paradox.

factualhigh valuecontestednovelty 3/4durability 3/4· Sean Carroll

“There seems to be, in Hawking's way of formulating the problem, no relationship between the information stored in the book that you throw into the black hole and the radiation that comes out other than how much mass the book had.”

0.66

It is not a law of nature that an information-bearing arrangement of particles weighs more than one without information; Go stones arranged to encode a message have the same mass as blank ones, though in some systems (like interacting spins) encoding information requires raising energy, so information-energy links depend heavily on specific assumptions and physical stuff is primary, with information merely carried by it.

factualhigh valuecontestednovelty 3/4durability 3/4· Sean Carroll

“it's not some strictly speaking law of nature that an information containing a collection of particles has more energy than one that does not contain information.”

0.63

Large language models like ChatGPT are not conscious or intelligent in any meaningful sense; they predict which words and sentence structures tend to appear together based on training text, so they have no concepts or world model and excel at sounding human while being unreliable about truth.

causalhigh valuecontestednovelty 2/4durability 2/4· Sean Carroll

“The large language model has zero understanding, it does not know what pizza is... because it doesn't know anything. Because it doesn't have any concepts in its brain”

0.57

Approval voting is unattractive despite advantages over plurality voting, because the binary approve/disapprove choice discards information about intensity of belief—someone you would go to the mat for and someone you can barely tolerate both receive the same approval—and intensity, not Arrow's theorem, is the more practically important concern in voting design.

normativehigh valuespeaker onlynovelty 3/4durability 3/4· Sean Carroll

“if I just vote for someone or don't vote for them, that's a harsh binary choice. It is not saying, 'Well, I really, really, really want this person and this person, I could take him or leave him.'”

0.57

It is an open research question what features of a low-level microscopic theory allow a higher-level coarse-grained emergent description to exist, and Carroll suspects such emergent descriptions are rare among possible microscopic theories, implying some non-naturalness to the existence of emergence, though he does not know the answer.

factualhigh valuespeaker onlynovelty 3/4durability 3/4· Sean Carroll

“even if all you do is ask, do most possible microscopic theories have an emergent higher level coarse-grained description, or is it very, very rare? I don't even know the answer to that. I suspect it's rare.”

0.56

Some shaping of which curiosity gets supported is unavoidable because research resources are finite and most ideas won't pan out, so advisors and institutions must judge which questions are worth pursuing; this can be done in good faith or can degenerate into curmudgeonly dismissal of anything unfamiliar, and the key is honestly distinguishing the two.

normativehigh valuespeaker onlynovelty 2/4durability 4/4· Sean Carroll

“in the real world, we have to choose, it's not a matter of policing so much as we have to choose to support investigation of some kinds of curiosity and not support others.”

0.56

When people say they don't have time to do something, they usually don't literally lack time but are not prioritizing it as highly as other things they want to do; the honest framing is to acknowledge the relative priority being assigned.

factualhigh valuespeaker onlynovelty 2/4durability 4/4· Sean Carroll

“when people say they don't have time to do something, they don't really mean they don't have time, they mean, I am not prioritizing it as high as I am the other things that I wanna do with my time.”

0.56

For naive utilitarianism, many-worlds and single-world-with-randomness give identical moral prescriptions when you weight utility by the wave function squared, but Lara Buchak's incorporation of risk aversion opens the possibility of a reasonable moral theory that genuinely distinguishes a 10% chance of something happening from it really happening to 10% of people (where there is a 100% chance it happens to someone).

factualhigh valuespeaker onlynovelty 4/4durability 3/4· Sean Carroll

“if you're not just maximizing utility, but you're taking risk aversion into consideration, then maybe you could imagine a reasonable moral theory, which distinguished between a 10% chance of something happening and a 90% chance of it not happening versus it really happens to 10% of the people”

0.56

Good questions arise because people are not empty vessels but bring pre-existing mental models, so questions emerge when a presentation leaves a gap in one's reconstructed understanding ('I expected you to say X but you didn't') or conflicts with one's prior view; asking good questions is a non-algorithmic art form improved mainly by practice.

factualhigh valuespeaker onlynovelty 2/4durability 4/4· Sean Carroll

“people are not empty vessels into which information is poured... People have pre-existing ideas about the universe. And when they're listening to something... you're trying to reconcile that with what you already know.”

0.56

Traits like humility and a growth mindset are hard to train directly—a student must already want them—but the most valuable thing a teacher can do is exemplify them: openly admitting mistakes and confusions while maintaining enough authority to be trusted, and letting genuine, unfakeable curiosity and passion show.

normativehigh valuespeaker onlynovelty 2/4durability 4/4· Sean Carroll

“the one thing I can say is, exemplify those virtues, if you want students to get them themselves. They're hard to train.”

0.52

The proposed 'prospector mindset' (wanting exclusivity and precedence in discoveries) is not a third category alongside Julia Galef's scout and soldier mindsets, because scout/soldier concerns one's attitude toward learning (open inquiry vs defending a position) whereas prospecting concerns what one does after learning—whether to hoard credit or share knowledge—and like capitalism it has both beneficial and detrimental aspects.

factualhigh valuespeaker onlynovelty 3/4durability 3/4· Sean Carroll

“you're not talking about that attitude that we have towards learning new things, you're talking about what we do once we learn new things. Do you want exclusivity and precedence, or do you just wanna share all of your knowledge widely? So that's a different distinction than scout and mindset.”

0.49

Simulating the universe requires an amount of computation equal to the computation the universe itself performs, which is very large; a smaller computer could in principle do it given enough time, but this is harder than it sounds because the universe interacts with itself, making it difficult to divide into independent pieces—especially in a quantum universe with entanglement.

factualestablishednovelty 2/4durability 3/4· Sean Carroll

“simulating the universe requires an amount of computation equal to the amount of computation being done by the universe, which is very large.”

0.49

The length of physics PhD programs (aspirationally around five years in the US) is not a real problem, and advisors have little incentive to keep theory students around since early students consume more time than they produce; the genuine concern is that we ask people to spend too long as postdoctoral scholars.

normativehigh valuespeaker onlynovelty 2/4durability 2/4· Sean Carroll

“I think that the issue is postdocs, honestly. The amount of time we take being a postdoctoral scholar, that's an issue that is weird and we have to face it.”

0.49

Claims that universities are taken over by 'woke' culture and that physics is in crisis with stifling groupthink are roughly false but exploit genuine critiquable issues; the people making them typically use emotionally charged language as ammunition for a political agenda (pushing right-wing views or promoting their own neglected physics ideas) rather than engaging the subtle real problems carefully.

factualhigh valuespeaker onlynovelty 2/4durability 2/4· Sean Carroll

“these legitimate issues are taken up as fuel or ammunition... for some political background viewpoint that they want to push.”

0.47

In the modern world there is a plausible case that individuals are both over-informed about distant events they cannot influence (causing useless fretting) and under-informed about important issues; personal well-being might improve with less news consumption, but a social obligation to be a good citizen means the right target is knowing just enough.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Sean Carroll

“we are sort of over-informed about some things just in the sense that we can fret about them and feel bad about them without having any power to do anything over them. At the same time... many people are under-informed about very important issues”

0.44

Simulated universes and real universes could be practically the same for any intelligent beings inhabiting them, with no necessary significant difference, a position argued at length in David Chalmers' book Reality+.

factualcontestednovelty 2/4durability 3/4· Sean Carroll

“I think they could be practically, the same. I don't think there's necessarily a significant difference”

0.44

We do not know whether infinity exists in the physical world; infinity is often a convenient approximation (e.g., calculus over a continuum is easier than discrete math), and we cannot currently distinguish truly infinite from very large, so we should keep an open mind rather than demanding immediate answers to such big questions.

factualcontestednovelty 2/4durability 3/4· Sean Carroll

“Infinity is a very useful hypothesis in many cases... So, it might be just a convenient approximation, but we don't know.”

0.44

Caltech uniquely excels at creating a low-bureaucracy space for brilliant science and engineering researchers to do focused work, but at the cost of cross-disciplinary interaction and a sense of being a comprehensive university, whereas institutions like Johns Hopkins make interdisciplinary collaboration and socializing across departments far easier.

factualhigh valuespeaker onlynovelty 2/4durability 2/4· Sean Carroll

“the thing that Caltech is good at is creating a space for brilliant creative people to go, especially sort of science and techy people to go and do their research in their area.”

0.42

There is no problem carrying a small black hole into a larger one; it can simultaneously be true that nothing entering the small hole's event horizon can escape to the surrounding region of the big hole, and that nothing in the big hole's region can escape to the outside world.

factualestablishednovelty 2/4durability 3/4· Sean Carroll

“there is no problem whatsoever in carrying a small black hole into a bigger one... Both of those can be true at the same time.”

0.34

Differences between places—the US east and west coasts, or even different countries—are much smaller than the similarities, because the same basic structures of human life (convenience stores, transport, farms) recur nearly everywhere, so apparent regional or national distinctiveness tends to be exaggerated.

factualspeaker onlynovelty 2/4durability 3/4· Sean Carroll

“what always strikes me about going to other countries is how similar things are. It's the same basic way of doing things.”

0.34

Knowing your genome is worthwhile because learning you are more susceptible to a serious disease is useful information that may let you prevent or mitigate it, and the main objection comes from third parties who distrust individuals to handle probabilistic risk; a real-world caveat is that many sequencing companies treat customers as data-providing products.

normativecontestednovelty 1/4durability 2/4· Sean Carroll

“to know that you are susceptible to getting a really bad disease more than the average amount seems like really, really useful information. Maybe you can take some steps to prevent it or mitigate it”

0.31

Professional sports likely make a very tiny contribution to societal violence and polarization, and the relationship may even be negative, since sports exist worldwide and rooting for a team can provide an outlet for natural aggressiveness, though there is no firm data either way.

causalspeaker onlynovelty 2/4durability 2/4· Sean Carroll

“rooting for a sports team gives a channel, an outlet for one's natural aggressiveness or desire to whoop it up, etcetera. So, I don't know of any actual data one way or the other”

0.24

Sleep advice is highly individual; if a recommended method to fall asleep doesn't work for you, you shouldn't be discouraged but should try others, because different people are just very different—Carroll's own solution was simply going to bed late enough to be very tired.

normativespeaker onlynovelty 1/4durability 2/4· Sean Carroll

“if you read about a way to help you fall asleep and you try it and it doesn't work, I wouldn't be too depressed by that or disappointed. I would just try other ways, I think different people are just different there.”