What this covers

Sean Carroll hosts a monthly question-and-answer episode drawn from Patreon supporter submissions, selecting questions primarily on whether he has substantive things to contribute rather than the merit of the questions themselves. The episode proceeds across physics, philosophy, media, and social topics, with Carroll defending a consistent intellectual stance: that Truth as Highest Value distinguishes genuine intellectual work from activism (which serves predetermined goals) or education (which transmits consensus). This frame organizes his treatment of quantum mechanics, cosmology, and moral reasoning, showing how tentative, mechanism-grounded thinking differs from both false certainty and destructive skepticism.

The episode ranges widely over quantum foundations and classical thermodynamics. Carroll argues that Bayesian Reasoning provides better guidance than crude testability, with decoherence selecting pointer states through identical interaction patterns rather than collapsing the wave function. He defends many-worlds interpretation as deriving structure cleanly, contests the misleading picture of virtual particles and the expansion of space within bound systems, and clarifies that quantum entanglement differs from classical entropy even where formulas coincide. On cosmology, he expresses skepticism about extrapolations to the very early universe while defending both general relativity and quantum field theory near black hole horizons—a joint consistency that yields the black hole information paradox.

Beyond physics, Carroll articulates why Fallibilism should remain live to brute facts—that some aspects of reality resist deeper explanation—while distinguishing between observer-dependent entropy (reflecting knowledge) and quasi-objective entropy (via coarse-graining). He contends that morality is constructed from inclinations plus rationality rather than discovered, that experimental philosophy confuses psychology with philosophy, that Steelmanning / Engaging Counterarguments signals trustworthy reasoning, and that caring about legacies makes sense precisely because you are alive now, even though you will not experience their fruits. Libertarian free will he dismisses as incompatible with any laws of physics.

Sharpest takeaway

Carroll argues that the highest intellectual value is the pursuit of truth itself — distinct from activism, education, or edginess — and applies this stance across physics (many-worlds, emergent spacetime, entropy) and philosophy (morality as construction, fallibilist Bayesianism) to show that good reasoning means holding tentative, mechanism-grounded views while keeping open the possibility that some facts are simply brute.

  • An intellectual seeks truth as the highest value, unlike activists (goal-driven) or educators (transmitting consensus)
  • Many-worlds is the cleanest interpretation because it derives rather than postulates structure, even at the cost of more work
  • Theories should be evaluated via Bayesian priors and updating, not crude testability, while keeping 'brute fact' as a live option

The claims · ranked57 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.

0.58

It is overwhelmingly (though never 100%) likely that the universe will 'end' in the sense that the arrow of time ends — the universe runs out of the fuel that lets entropy keep increasing — rather than time itself ending; this idea (the heat death) leaked into popular culture in the late 19th/early 20th century (Henry Adams, arguably Nietzsche's eternal recurrence, Thomas Pynchon) but has limited everyday impact because the relevant timescales are so vast that people cannot relate them to their lives.

factualestablishednovelty 2/4durability 4/4· Sean Carroll

“it's not that time will end, but the arrow of time will end, we'll use up the fuel that we have to keep the universe going.”

0.58

Morality is a systematization and codification of our deep-seated moral inclinations (feelings that things are right or wrong) combined with rationality (the drive to make those feelings into consistent principles); because inclinations can be inconsistent and people differ, morality is not objective, universal, or cut-and-dried, and different people may construct different moralities — yet Carroll is not a moral relativist, holding it is valid to apply your constructed rules to others. He applies moral rules species-by-species rather than individual-by-individual, calling the 'name the trait' argument's demand for individual-level application arbitrary.

definitioncontestednovelty 3/4durability 4/4· Sean Carroll

“what we call morality is sort of a systematisation and codification of our moral inclinations, of our deep seated feelings, that certain things feel right, certain things feel wrong, plus, our rationality”

0.58

An intellectual is defined not by intelligence but by treating the pursuit of truth as the highest value, actively working at the boundary of what is known — distinct from activists (whose primary loyalty is a goal that can conflict with truth-telling) and educators (who transmit existing consensus); valorizing contrarianism, edginess, or heresy is fundamentally anti-intellectual because it valorizes a posture rather than the truth, even though being public creates a tension since audience-building strategies may not align with truth-telling.

definitioncontestednovelty 3/4durability 4/4· Sean Carroll

“An intellectual is someone who is characterised by... someone who seeks truth as the highest value.”

0.58

Despite much bad post-modernism, there is a valuable core insight: even if an objective world exists, we are tiny, finite, biased beings who never know it in an unmediated way, so all our stories about the world — even rigorous physics theories — are partial and fallible; this is trivial for physics (apples still fall) but crucial in the social, ethical, and political domains, where it is far easier to be fooled into thinking you have figured things out while embedded in unquestioned presuppositions, and post-modernism provides a toolkit for questioning those presuppositions.

factualcontestednovelty 3/4durability 4/4· Sean Carroll

“even if the world exists and is objective, we don't know it clearly and objectively, and in an unmediated way. We are tiny, finite, biased beings”

0.54

The second law of thermodynamics and entropy are not fundamental — Laplace's demon, knowing the exact microstate, would have no ignorance and thus no entropy — but they are real in the sense that entropy is a higher-level emergent concept with genuine causal and explanatory power; this exemplifies Carroll's general view that there are better and worse ways of dividing the world into objects, where a collection counts as real if it has a causal role and an autonomous higher-level theory, unlike an arbitrary collection (your foot plus a coffee cup plus an oak tree).

factualcontestednovelty 3/4durability 4/4· Sean Carroll

“the fact that there is a useful concept called entropy, that has causal explanatory power... means that there is some reality to that concept, so it's not fundamental, but it is real.”

0.54

Given that not all possible universes exist, some fundamental features of our world are explained by nothing deeper than 'that's the way it is' (brute facts), but we can never know which features are brute and which have deeper explanations, nor can we ever know we have reached the bottom level of reality; the right attitude is to keep the 'brute fact' option live while still searching for deeper explanations, never insisting there must be a satisfying answer — because science never reaches a point of complete certainty.

factualcontestednovelty 3/4durability 4/4· Sean Carroll

“There are some facts about the universe that are explained by nothing deeper than, 'That's the way it is,' brute fact, cold, hard fact. But we don't know which features are like that and which are not.”

0.54

Rather than the crude binary of testable-versus-untestable, theories should be evaluated as Bayesians: people assign priors based on values like simplicity, fruitfulness, and comprehensiveness (you are entitled to your own priors but not your own likelihoods), then update on evidence — which can take many forms beyond experimental data, including new theoretical insights or the realization that a theory is more complicated or vaguer than thought, all of which count as new evidence.

normativecontestednovelty 3/4durability 4/4· Sean Carroll

“We should be Bayesians, we should say, 'We assign priors to different hypotheses about the universe based on whatever criteria it is that we use to assign priors'”

0.54

Skeptical scenarios (Descartes' demon, Boltzmann brain, simulation) are genuinely possible and so convincing that no experiment could rule them out, but a good Bayesian discounts them by assigning very low — though never exactly zero — prior credence, because excluding the Boltzmann brain possibility keeps us in a cognitively stable situation: if you grant you might be a Boltzmann brain, you have no reason to trust any of your experiences and cannot make sense of the world; credences must stay non-zero so future evidence (e.g., a demon revealing itself) could update them.

normativecontestednovelty 3/4durability 4/4· Sean Carroll

“to exclude the Boltzmann brain possibility by just setting the prior credence on it to a very, very, very low number, keeps us in a cognitively stable situation.”

0.54

We should be more humble about extrapolating established physics into untested regimes: extrapolations to black holes are very likely fine because nothing breaks down except in identifiable ways, but extrapolations to the very early universe (inflation, baryogenesis) are far less trustworthy because they rely not on robust features of quantum field theory but on specific features of specific quantum field theories — in which Carroll has little faith when pushed to the early universe; the domain of applicability of a theory is a question both physicists and philosophers should think harder about.

normativecontestednovelty 3/4durability 4/4· Sean Carroll

“they're often not simply relying on robust features of quantum field theory, but they're relying on specific features of specific quantum field theories. And those, I don't have much faith in at all, when we're extrapolating them to the very, very early universe.”

0.54

Whatever is going on in consciousness, you should not try to explain it by changing the laws of physics, because everything we know about effective quantum field theory says electrons interact purely locally — the only thing affecting an electron is the values and derivatives of other fields at its exact location — so an electron 'could not possibly care less' whether it is at CERN or embedded in a human brain (where particles are actually far apart by particle-physics standards); there is zero expectation of dynamics changing in complex systems, though it cannot be proven absolutely impossible.

causalcontestednovelty 3/4durability 4/4· Sean Carroll

“everything we know about the rules of effective quantum field theory, says the electron could not possibly care less, whether is embedded in a human brain. Electrons interact locally.”

0.54

Evaluating whose opinions to trust on technical topics is a matter of heuristics, not algorithms: weight world-renowned experts in precisely the relevant field more, but go beyond credentials to track record (papers, ideas that proved valid); discount expertise applied outside its domain; don't entirely dismiss non-credentialed people who may be insightful; and crucially, weigh how someone argues — whether they show they understand and engage the counterarguments rather than strawmanning or being snarky — as a strong signal of how seriously to take them.

normativeestablishednovelty 2/4durability 4/4· Sean Carroll

“what is the evidence that they're giving you, that they appreciate why they might be wrong, that they understand the counter arguments to it”

0.54

Finding galaxies that lack dark matter is counterintuitively evidence FOR the dark matter hypothesis, because the leading alternative — modified gravity — predicts that all similar collections of visible matter must produce the same gravitational field, whereas dark matter introduces an extra degree of freedom that allows two equal-sized collections of stars to differ, one with associated dark matter and one without.

causalestablishednovelty 4/4durability 3/4· Sean Carroll

“the counter-intuitive fact that finding evidence for galaxies without dark matter in them is evidence for dark matter.”

0.53

The picture of the quantum vacuum as particles popping in and out of existence is deeply misleading — a remnant of insisting on a visualizable classical world rather than quantum states of fields — but it does capture real complicated structure: a real particle traveling through the vacuum can equivalently be described either as moving through a pre-existing cloud of virtual particles and interacting with them, or as spitting out and reabsorbing virtual particles, both being perfectly good ways of thinking about it because physicists are constantly translating math into imperfect words.

definitionestablishednovelty 3/4durability 4/4· Sean Carroll

“the whole picture of the quantum vacuum as consisting of these particles popping in and out of existence is deeply misleading, still a remnant of our insistence on thinking of the world as something visualisable and classical rather than quantum states of fields”

0.53

Quantum (von Neumann / entanglement) entropy is a distinct form of entropy from classical Boltzmann entropy, even though their formulas coincide; quantum decoherence necessarily increases a system's entropy as it becomes entangled with the environment, but the entropy increase when mixing cream into coffee or shuffling cards is purely classical and has nothing to do with quantum entanglement — so it is incorrect to claim that growth of quantum entanglement drives the macroscopic arrow of time.

factualestablishednovelty 3/4durability 4/4· Sean Carroll

“Quantum decoherence necessarily increases the entropy of a system as long as you don't collapse the wave function and reset”

0.53

There are multiple definitions of entropy, but a key distinction is between an entropy that characterizes an observer's knowledge (a probability distribution over microstates, via Boltzmann/Gibbs) and a quasi-objective entropy obtained by first subjectively choosing observable macrostates (coarse-graining microstates into observationally indistinguishable groups) and then objectively assigning entropy as the logarithm of a macrostate's size; the two coincide when the observer's knowledge is precisely that the system is in a given macrostate and nothing more.

definitionestablishednovelty 3/4durability 4/4· Sean Carroll

“The subjective step was, you defined macrostates, the objective part is once you define that, there's an entropy that you can associate with every microstate.”

0.50

Surveying people's intuitions about free will or morality (experimental philosophy) is learning about psychology, not philosophy, because philosophical ideas stand or fall on the reasons for them, not on a poll or popularity contest; what the majority of people have in mind by a term like 'free will' is not an important consideration — the goal is the most useful vocabulary to describe the world.

normativecontestednovelty 3/4durability 3/4· Sean Carroll

“I don't think it's philosophy. I think it's an interesting thing to do, but it's more like you're learning about psychology, you're not learning about philosophy.”

0.49

The single biggest benefit of being at a top research university is being embedded in a network of smart people who both generate ideas and act as reality checks; while the internet allows remote participation in seminars, the most valuable interactions happen informally at lunch and coffee, which is why physical co-location still matters even though much can now be ameliorated online.

factualestablishednovelty 2/4durability 3/4· Sean Carroll

“the most important thing is that you're in a network of people who can both give you ideas and also act as reality checks on the ideas that you have.”

0.49

If Feynman hadn't invented his diagrams, physics would look much the same because someone else would have invented them (they are too useful), and even absent diagrams entirely, Julian Schwinger showed the same QED calculations could be done purely algebraically — so the calculations would have been done but harder, slower, and with less conceptual clarity, though perhaps a different clarity freed from visualizing particles literally moving on lines when it is really quantum fields involved.

factualestablishednovelty 2/4durability 3/4· Sean Carroll

“Julian Schwinger, who shared the Nobel Prize with Feynman and Tomonaga, basically did the same calculations that Feynman did, without the diagrams.”

0.49

Libertarian free will — the idea that choices made in our brains are not subject to the laws of physics — is a silly and arguably undefinable notion; if a human were found behaving in violation of physical law, the right response would be to revise the laws of physics to include that behavior, not to declare lawlessness. Libertarian free will is therefore mutually exclusive with any laws of physics (probabilistic or not), and most conceptions of free will are incompatible with a block universe, which classically goes hand-in-hand with determinism (you cannot be 'set in probabilistic stone').

factualcontestednovelty 2/4durability 4/4· Sean Carroll

“if a human being was found to be behaving in some way that was clearly in violation of the laws of physics, what I would do is change my idea of what the laws of physics were”

0.49

It is simultaneously true that some physics problems (the measurement problem, the role of probability in cosmology, the origin of the arrow of time) genuinely benefit from philosophy, and that most of physics and most of philosophy are entirely separate yet each valid in its own right — you don't call a philosopher to calculate a Feynman diagram, and ethics or aesthetics have no impact on physics; philosophy is interesting because philosophy is interesting, just as physics is interesting because physics is interesting.

normativecontestednovelty 2/4durability 4/4· Sean Carroll

“there are plenty of physics problems for which philosophy is not useful, that's just obvious, and so there's no like no necessary connection between physics and philosophy. If you wanna calculate a Feynman diagram, you don't call a philosopher.”

0.49

Since there is no afterlife, a legacy is useless to you after death — but this does not make caring about a legacy irrelevant, because you can validly take pleasure now in establishing something that pays off after you are dead; conversely, 'this is the one life so live it as you please' is a non sequitur smuggling in extra premises — the correct conclusion is that the experiences you have and the impact you have on others, here and now, are what matter, going far beyond your brain's whims and fancies.

normativecontestednovelty 2/4durability 4/4· Sean Carroll

“to take pleasure in establishing something now, that will only pay off after you're dead, there's nothing wrong with that. Even though you won't be around to appreciate it, you're appreciating it now”

0.48

Pointer states — the definite outcomes that survive decoherence (like a dot at one point on a detector screen, or an alive-vs-dead cat) — are selected because they are the parts of the wave function whose components all interact identically with the rest of the world, so they don't keep gaining entanglement; an electron hitting at a single point interacts the same way (emitting photons identically), whereas a smeared circle would have parts interacting differently (photons from different angles), causing those parts to decohere from each other.

causalestablishednovelty 3/4durability 3/4· Sean Carroll

“what pointer states are, are parts of the wave function which have the property that once they're there, once you're in them, they don't keep getting extra entanglement with the rest of the world.”

0.47

Contrary to the assumption that black holes are exactly where general relativity and quantum field theory break down, both should be trustworthy near a large black hole's event horizon, because GR fails only where curvature blows up (the singularity, not the horizon, where curvature can be small) and QFT is trustworthy near the vacuum with few particles (which holds near the horizon); the genuine puzzle is that both being trustworthy there jointly produces the information loss paradox, suggesting one (probably QFT, via some subtle non-locality) is wrong near the horizon.

factualcontestednovelty 3/4durability 4/4· Sean Carroll

“near the event horizon, the curvature of space time can be small, especially for a big black hole. So that's precisely a regime in which you should trust general relativity perfectly.”

0.47

It is an empirical fact that the initial state of our observable universe was special — describable as low entropy or maximally unentangled — and why this is so remains a major unanswered question in cosmology; Carroll proposed a baby-universe model with Jennifer Chen as a favorite candidate explanation, but it remains open.

factualestablishednovelty 2/4durability 4/4· Sean Carroll

“that's a major unanswered question in cosmology, why did the early universe had a special state which can be described as low entropy or maximally entangled?”

0.47

Electrons are used as the standard quantum mechanics example for two reasons: they are elementary (not made of other particles like protons and neutrons, which are made of quarks and gluons), avoiding internal-structure complications; and they are the lightest charged particles, so since electromagnetism drives the everyday world, electrons do most of the work (electricity is mostly electron motion because moving electrons left requires far less energy than moving protons right) — but everything, including protons and neutrons, has a wave function.

factualestablishednovelty 2/4durability 4/4· Sean Carroll

“electrons are the lightest charged particles in nature... since electromagnetism is really the force of nature that makes everyday world go, electrons are doing most of the work”

0.47

Although bizarre events do occur in some branches of the wave function, you cannot move qualitatively from 'some unlikely things happen' to 'they probably explain anomalies,' because the probabilities must be quantified: macroscopic quantum events (like a coffee cup tunneling through a table) have probabilities around 10 to the minus a million, vastly smaller than the ~10^20 opportunities for them to occur, so it is far easier to explain ghost stories by misperception than by many-worlds.

causalestablishednovelty 2/4durability 4/4· Sean Carroll

“the chances that the coffee cup tunnels through the table is 10 to the minus a million, which is a very, which actually overestimates the probability”

0.47

While the ability of one mechanism to explain many things is rightfully prized in an explanation (per Simon DeDeo), making people wary when the right answer involves disconnected unrelated things, the real world very often does require invoking disconnected unrelated causes — as with element formation, where the Big Bang produces helium but stars and supernovae are needed for heavier elements, vindicating both Gamow and Hoyle in part.

factualestablishednovelty 2/4durability 4/4· Sean Carroll

“the ability of one thing to explain many things, is rightfully thought of as a good feature for an explanation to have, and this makes people a little bit leery... when the right explanation seems to involve disconnected unrelated things. But that's really very often true in the real world”

0.47

Mach's principle — that the distant distribution of mass sets the rotating-versus-non-rotating reference frame or that distant matter determines local inertia — is simply wrong; it inspired Einstein toward general relativity's general idea that matter-energy relates to spacetime curvature, but the specific Machian claim is not part of general relativity, which has matterless solutions that still distinguish rotating from non-rotating, so once you have general relativity you no longer need Mach's principle.

factualestablishednovelty 2/4durability 4/4· Sean Carroll

“Mach's principle is wrong. That's the answer.”

0.47

You cannot undo the past by making quantum measurements (e.g., running a universe-splitter app many times); if you ask a priority question today, then all future versions of you descend from someone who asked it today — quantum branching does not let you escape onto a 'different timeline' that erases prior events.

factualestablishednovelty 2/4durability 4/4· Sean Carroll

“You can't undo the past by making measurements with the universe splitter or any other quantum mechanical measurement process.”

0.47

The claim that 'space is expanding' is an exaggeration; only parts of space expand, and it is impossible to attach definite ideas of where space is expanding — within a gravitationally bound system like a galaxy, planet, or person, there is no expansion at all; what is true is that distant galaxies are farther away than before, so it is only the regions between galaxies that actually expand.

factualestablishednovelty 2/4durability 4/4· Sean Carroll

“within a gravitationally-bound system, like a galaxy or a planet or a person, there's no expansion of space going on; it's only the regions in between the galaxies that are actually expanding.”

0.44

Pascal's wager fails for two reasons: trivially, the same infinite-reward reasoning could justify contradictory propositions (e.g., a God who damns you for believing for Pascalian reasons), so it gives no guidance on which religion to choose; and more deeply, multiplying a small prior by infinity to force a decision is itself a reason not to take such infinities seriously, since humans cannot meaningfully distinguish infinity from a very large number.

factualcontestednovelty 2/4durability 3/4· Sean Carroll

“There's a chance that God exists and would be really annoyed if you believed in Him, or worshipped Him for the wrong reasons... So just out of the gate, I think that Pascal's wager is not a very good argument at all.”

0.44

Just as Carroll is not a moral realist, he is not an aesthetic realist: there is no objective superiority of different works of art; what matters is the relationship between the artwork and the experiencer, so the most useful criticism describes the experience non-judgmentally ('if you like X you'll like this') rather than asserting objective quality, and reviews that downgrade a work for not being the genre the reviewer wanted are a category error.

normativecontestednovelty 2/4durability 3/4· Sean Carroll

“I don't believe in the objective superiority of different works of art, I think that the important thing about art is its relationship to the experiencer of the art.”

0.44

The accusation of being 'post-truth' is almost always used as a slander against people you dislike rather than describing anyone's actual self-understanding; almost no sensible person argues truth is overrated, so the right move is to confront opponents honestly on where their evidence and reasoning go wrong rather than claiming they aren't aiming at truth at all — which is itself an activist rather than intellectual move.

normativecontestednovelty 2/4durability 3/4· Sean Carroll

“the phrase post-truth is almost always used as a slander against people you don't like.”

0.44

A nation of millions of people is tentatively not conscious; while integrated information theory (IIT) attempts to quantify a level of emergence at which something becomes conscious — allowing for the possibility that the locus of consciousness could attach to a coordinated mob rather than its individuals — this framework is not helpful for a nation, and although we routinely attribute beliefs and psychologies to countries, that is shorthand for what should be more nuanced talk.

factualcontestednovelty 2/4durability 3/4· Sean Carroll

“We do it all the time, we do the equivalent thing because we attach beliefs, desires, attitudes, psychologies to countries of people, but it's often just a short hand for what should be a more nuanced way of talking about things.”

0.43

The role of the algebra of observables is less clear in many-worlds than in other interpretations precisely because other interpretations add it by hand as extra superstructure, whereas many-worlds — the simplest, cleanest interpretation — must derive observables, branching, and probabilities as emergent; this is simultaneously an advantage (deriving rather than positing) and a disadvantage (it is more work and you might be fooling yourself that the derivation succeeds), and Carroll's quantum mereology work with Ashmeet Singh tackles exactly this.

factualcontestednovelty 3/4durability 3/4· Sean Carroll

“in all of the other interpretations, you put the algebra in there by hand, you add extra super structure, whereas many worlds is the simplest, cleanest interpretation of quantum mechanics. So when it comes to things like observables... you have to derive them, they're emergent.”

0.43

For wormholes and Alcubierre warp drives, effective faster-than-light travel works by warping spacetime locally (never exceeding light speed locally), and how much time elapses on the traveler's clock depends on the details of the configuration; it is conceivable to set up a wormhole so that the elapsed traveler-time equals the elapsed home-time, meaning a Star Trek/Star Wars-style FTL with no relativistic mismatch is conceivable — though it is not what naturally pops out of realistic spacetime geometry.

factualcontestednovelty 3/4durability 3/4· Sean Carroll

“I could easily imagine setting up my completely fictitious worm hole system, such that the time that elapsed on my watch, as I travel through the warm hole and travel back, is the same amount of time as elapsed to the people back home.”

0.43

Inflation is touted for solving the horizon problem by smoothing the early universe, but Carroll's objection is that we don't know why inflation ever began; whatever caused inflation to start might instead have just produced a big smooth universe directly — and the real problem is that the early universe was finely tuned and smooth, with our currently observable universe being only about a centimeter across when the expansion rate was at the Planck scale, an enormously large size by particle-physics standards; Carroll has no second-favorite alternative to inflation for solving this.

factualcontestednovelty 3/4durability 3/4· Sean Carroll

“my problem with inflation is I don't know why it ever began in the first place. But whatever thing caused inflation may be to begin... maybe you could have skipped the step of leading to inflation and just given us a big, smooth universe.”

0.43

The Reeh-Schlieder theorem establishes that any two regions of the quantum vacuum share some entanglement and that projecting on one region can in principle turn the rest of the universe's wave function into nearly anything — but this does not resolve the black hole information paradox, because (like the impossibility of choosing your spin outcome to enable superluminal signaling) you cannot force or choose these projections, the relevant operations are extraordinarily unlikely, and mere existence of entanglement across the horizon is not enough to get the information out.

factualcontestednovelty 3/4durability 3/4· Sean Carroll

“The Reeh-Schlieder theorem implies that there is entanglement between degrees of freedom inside the black hole, and degrees of freedom outside the black hole. That's true, but that doesn't help you get information out.”

0.43

There is no such thing as one model of national government 'winning out' as optimal; the global equilibrium will continue to feature many models (democracies, authoritarian, socialist). What matters about democracy is not merely that it may be a better problem-solver than authoritarian systems (per Henry Farrell) but that it puts power in individuals' hands, unleashing creativity — the same value of allowing people to be rebellious and go against the grain that drives scientific progress, putting authoritarian societies at a disadvantage in a world where scientific and technological progress matter.

forecastcontestednovelty 2/4durability 2/4· Sean Carroll

“What matters to me about democracy is that it puts power in the hand of individual people, and that power helps unleash their creativity.”

0.43

Fred Hoyle is perceived as a 'rated genius' rather than a crackpot; his stubborn insistence on the steady state universe was quirky and cost him some favor, but originators of ideas get some slack for clinging to them, and everyone agrees he made great contributions to theoretical physics and astrophysics, especially nucleosynthesis and the origin of the elements.

factualestablishednovelty 1/4durability 3/4· Sean Carroll

“everyone agrees that Fred Hoyle did a lot of great things in Theoretical Physics and Astrophysics, especially nucleosynthesis and the origin of the elements.”

0.42

One can imagine knowing the complete microstate of a system without knowing how to chunk it into emergent higher-level concepts — analogous to being handed an operating system printed as raw binary without being told what it is: all the information is present, but a separate layer of conceptual insight is required to recognize it, so a microstate-level Laplace's demon does not automatically entail a macrostate-level understanding.

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

“if someone takes... an operating system... and translates it into binary... and hands it to you without telling you what it is. Could you figure out what it is? I'm predicting, the answer is no.”

0.41

The recent hostility of physicists to philosophy may correlate with physics' center shifting from Europe to the United States, bringing a pragmatic, get-it-done attitude that, combined with a pragmatist view of truth as the 'cash value' of ideas, points physicists toward ignoring philosophical questions that don't help predict experimental outcomes; compounding this is the hyper-specialization and 'siloisation' of the academy, where departments self-perpetuate (hiring done ~98% within existing departments), devaluing cross-disciplinary perspectives.

causalspeaker onlynovelty 3/4durability 3/4· Sean Carroll

“if your bigger picture purpose is predicting the outcomes of physics experiments and your view of truth is essentially pragmatic, what is the cash value of these ideas, then you can see how you would be pointed in the direction of ignoring philosophical questions”

0.41

The cosmological multiverse (classical, with distant regions having different local physics) and the Everettian many-worlds (wave-function branching) are conceptually distinct, but they can be connected: if the branching events are themselves transitions between vacuum states in a landscape (as in string theory), then different branches can effectively contain different parts of the cosmological multiverse — an idea pursued by Yasunori Nomura and by Susskind/Bousso — though it didn't take the field by storm because it doesn't obviously solve any problem like the low-entropy early universe.

factualfringenovelty 3/4durability 3/4· Sean Carroll

“you could sort of combine the many worlds idea with the cosmological multiverse and say that in different branches you had effectively different parts of the cosmological multiverse.”

0.39

Going against conventional wisdom should not be valorized for its own sake; being a heretic is hard work and only justified in individual cases where one has genuinely concluded the conventional view is wrong, because there is a romantic attraction to heresy — both personally and for building an audience — that is dangerous to the value of truth-seeking.

normativespeaker onlynovelty 3/4durability 3/4· Sean Carroll

“being a heretic is hard work... it's not good to just be a heretic for the sake of being a heretic, you have to be a heretic because there is some individual case where you think it's worth going against the conventional wisdom”

0.39

Our ability to predict real scientific and technological advances is uneven because the easy-to-predict advances are extrapolations of things we already do (faster travel, going further), whereas conceptually new categories (like the internet) are hard to imagine; speculative ideas that naturally suggest themselves to humans (like superluminal tunnels) are drawn from extending current capabilities and are largely disconnected from the space of ideas that could actually be true.

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

“the kinds of advances that it's easy to predict are taking things we already do and imagine we can do them better... whereas a conceptually new category like the Internet, is harder to imagine.”

0.39

Carroll uses David Lewis's space of all possible worlds as a tool for thinking through self-locating uncertainty about the laws of physics, but he is not a modal realist — he believes in the reality of Everettian branches of the wave function but not in the reality of all possible worlds; so the question of why these laws of physics hold rather than others does not reduce to self-locating uncertainty among actually existing worlds (a brute fact), which is why the parallelism between modal realism and Everettian quantum mechanics breaks down for him.

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

“I'm not a modal realist, I am an Everettian, and I do believe in the different branches of the wave function, but I don't believe in the reality of all the possible worlds.”

0.39

Many-worlds has no moral implications under ordinary or utilitarian ethics, where a quantum gamble (e.g., spin-up gives everyone $1, spin-down gives everyone $2) is fine because each branch is good; it could matter only under contrived non-linear moral codes — for example, if your highest value were equality across the multiverse, you would refuse the bet so that no branch has people getting different amounts. For the self-sacrifice case, a branch where everyone lives plus an equal branch where everyone dies is, for Carroll, no different from a 50-50 probability in a single world, because residents of each branch experience exactly what single-world residents would.

normativespeaker onlynovelty 3/4durability 3/4· Sean Carroll

“if for some reason, you had this cockamamie moral theory that says that inequality is intrinsically bad even across the multiverse, then you would say it's bad that in some part of the multiverse people get $1 and in the other part they get $2”

0.38

A major triumph of 20th century physics is that space and time are two aspects of one thing (spacetime), so proposing emergent-spacetime theories that treat space and time differently — where time is fundamental but space is not — is in a real sense a step backward toward Newton, and the loss of that achieved progress is a legitimate strike against the approach; if such a theory is right, then the apparent unity of space and time must be an accident that emerges at low energies, which the theory would have to explain.

normativefringenovelty 3/4durability 3/4· Sean Carroll

“in some very real sense, it's a step backwards. And you can ask, 'Is that a strike against this approach?' And I think the answer is yes.”

0.37

Theories like GRW, in which wave functions spontaneously collapse with a built-in arrow of time, tuned to be frequent enough to explain macroscopic experience yet infrequent enough not to show up in existing experiments, ask too much and are very awkward; many-worlds offers a more elegant explanation already on the table, though Carroll would not literally retire from physics if GRW turned out true.

normativecontestednovelty 2/4durability 3/4· Sean Carroll

“they collapse in one direction of time, but they don't anti-collapse and it's exactly tuned to be frequent enough to explain our experience of the macroscopic world. But infrequent enough not to show up in experiments that we've already done, that just seems to be asking a lot”

0.37

If space emerges from quantum entanglement of abstract degrees of freedom, then the expansion of the universe can be reconceived not as new space being created (the classical view) but as more previously-unentangled degrees of freedom becoming entangled into the existing spacetime; the Planck scale is not a pixel-size resolution but a measure indicating where gravity, relativity, and quantum mechanics must all be jointly accounted for, and it does not increase as the universe expands.

factualfringenovelty 3/4durability 2/4· Sean Carroll

“what you mean by the expansion of the universe is more and more of these unentangled degrees of freedom are becoming entangled and space appears to expand.”

0.35

In the United States, the single most effective individual action to fight climate change is to get Democrats elected, because strong political polarization means Democrats broadly favor action while Republicans broadly do not; climate problems require systematic solutions implemented through governments, not valorous individual choices like recycling or driving an electric car, and once a willing government is in place you can still pressure it to do more.

normativecontestednovelty 2/4durability 1/4· Sean Carroll

“the single most effective thing you can do to help fight climate change is get Democrats elected into office.”

0.34

Life based on nuclear physics on a neutron star (as in Robert Forward's Dragon's Egg) is most likely incompatible with known nuclear physics, because chemistry is interesting precisely because atoms can be hooked together in arbitrarily complicated ways — DNA being an aperiodic crystal (as Schrödinger predicted) that can be arbitrarily long without repeating, enabling information storage — whereas things made only of neutrons or neutrons-and-protons cannot achieve arbitrary complexity.

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

“the great thing about DNA is, it's an aperiodic crystal, as Erwin Schrödinger predicted even before it was discovered, which means it can be as long as you want without ever repeating itself.”

0.34

Physical space and distance may be emergent from the entanglement between abstract quantum degrees of freedom — not entanglement between macroscopic objects like people, but between invisible degrees of freedom in the space between them — where specific geometric patterns of entanglement knit together into a smooth spacetime manifold; however, whether these entanglement structures merely resemble distances or actually constitute them requires a dynamical theory linking entanglement to signal propagation speed, which is still unresolved.

factualfringenovelty 3/4durability 2/4· Sean Carroll

“the different amounts of entanglement knit together to form a more or less smooth three-dimensional spacetime.”

0.32

Carroll's current Bayesian assessment is that the Fermilab muon g-2 result is less likely to be new physics than not — not because of the experiment (which cleanly measured the muon's magnetic moment) but because of priors: many claimed Standard Model anomalies have failed to pan out, and crucially the discrepancy depends on a hard theoretical prediction involving strongly interacting quarks and gluons, where one calculation yields the discrepancy but a lattice QCD calculation actually agrees with experiment.

forecastcontestednovelty 2/4durability 1/4· Sean Carroll

“there's another calculation done with lattice QCD, which actually agrees with the experimental result. So if you're looking for a way that the result might actually not be new physics, it's pretty obvious what it would be, that the theory prediction is just a little bit off.”

0.30

Most History/Science/Discovery Channel science shows are made poorly because TV networks hire production companies whose writers and producers — none of them scientists — research topics via Wikipedia and YouTube and decide content, interviewing actual scientists only on camera; the individual segments may be respectable but the overall vision lacks scientific guidance, leading to embarrassing losses of meaning across the many production steps (e.g. a dark matter demonstration whose analogy nobody on set understood).

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

“none of the people involved in that are scientists. They interview the scientist once they get on camera, but the writing and the organising, and deciding what to do on the show, is rarely guided by people who are knowledgeable about science.”

0.20

Carroll's podcast production process is not magic: he finds guests via suggestions and his own interests, records via Zencastr, edits lightly in Audacity (removing coughs, leaving ~99% intact), and the whole recording-editing-publishing cycle takes about one day a week, plus variable research time to prepare for less-familiar guests.

factualspeaker onlynovelty 1/4durability 1/4· Sean Carroll

“the whole recording, organising, editing, publishing, takes a day, roughly speaking.”

0.13

Shorter questions are more likely to get answered in the AMA, not because long questions are worse, but because a short question lets Carroll quickly recognize he can answer it in a brief period; it's necessary but not sufficient.

factualspeaker onlynovelty 0/4durability 1/4· Sean Carroll

“if you wanna get your question answered, shorter questions definitely do help you have that happen.”