AMA | November 2021
What this covers
This November 2021 Ask Me Anything episode presents Sean Carroll answering Patreon-submitted questions across science, philosophy, technology, and other domains. The conversation is loosely structured around broad clusters of inquiry rather than a single narrative thread, allowing Carroll to move between specific technical claims and wider methodological positions. The backbone is Bayesian credence reasoning—the practice of assigning rational confidence levels to claims and updating them with evidence—applied across everything from black hole physics to consciousness studies to cosmology.
The episode spans several distinct territories. On quantum mechanics, Carroll clarifies common misconceptions: the Heisenberg uncertainty principle as a statement about Hilbert space geometry rather than measurement disturbance, and the true nature of Albert Einstein's objection to quantum theory (locality concerns in the EPR paper, not merely indeterminism). On cosmology, he defends inflation as the leading paradigm while entertaining pre-Big-Bang scenarios that avoid treating the Big Bang as special. He addresses Selection effect / Anthropic reasoning applied to the cosmological constant, arguing it remains tentative absent confirmation of a multiverse. On consciousness and identity, Carroll argues these are Coarse-graining / Emergence phenomena grounded in established physics, not separate ontology—and that exotic alternatives like telepathy or consciousness-driven reality fail on grounds of both credibility and precision compared to the core theory. More broadly, he defends why Fallibilism does not require assigning low credence to all current theories, and emphasizes the principle that Never assign zero credence because it locks reasoning against future evidence.
Carroll argues that the established laws of physics (the 'core theory') are robust enough that exotic explanations—telepathy, panpsychism, consciousness as fundamental, modified redshift—should receive vanishingly low credence, and that most apparent puzzles about consciousness, identity, and cosmology dissolve under careful poetic-naturalist analysis rather than requiring new fundamental ingredients.
- Higher-level concepts like personal identity and consciousness are useful emergent descriptions of underlying physics, not illusions and not separate ontology
- Extraordinary claims that would overturn well-tested physics deserve far lower credence than mundane alternatives
- Scientific reasoning is fallibilistic and ultimately empirical, but theoretical insight can also legitimately shift credences
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.
Scientific thinking is fallibilistic and ultimately empirical, but theoretical insight alone—without any new experiment—can dramatically boost or squelch credence in a theory, as demonstrated when 't Hooft proved spontaneously broken gauge theories are renormalizable in the early 1970s, causing interest in Weinberg's previously ignored 1967 electroweak model to surge from near-zero to becoming the most-cited particle physics paper.
“So the only thing that changed in 1972 was a theoretical insight.”
One should never assign exactly zero credence to any logically consistent possibility, because in Bayes's formula multiplying a zero prior by any finite likelihood always yields zero, so no future evidence could ever change your mind—every possibility should retain some small nonzero credence to remain open to updating.
“if you give zero credence to something, then literally no future information can make you believe it, right? You're multiplying zero by some finite number, you're still gonna get zero in Bayes's formula.”
Theoretical physicists propose a theory not because they think it is true but because they think it is possibly true and worth investigating, assigning it some credence; Carroll's most-cited paper proposes modified gravity to explain cosmic acceleration, yet he thinks the cosmological constant is the more likely explanation, illustrating that proposing a model and believing it are distinct.
“You propose an idea, not because you think it's true, but because you think it's possibly true, and you assign some credence to it.”
Cosmological scenarios dominated by random fluctuations should be assigned very low credence not because we are 'ruled out' as Boltzmann brains by data, but because such scenarios imply that essentially all of one's beliefs about the external world—including the reasoning that led to the scenario itself—are randomly fluctuated and false, making the theory self-undermining; yet one should not assign exactly zero credence, since zero credence is impervious to all future evidence.
“The real problem with the Boltzmann brains narrative is not that it's ruled out by data, but that it leads you to conclude that all of your beliefs about the world are wrong, including the ones that led you to believe that there's a Boltzmann brain problem.”
When evaluating a surprising claim like rats worldwide learning tasks faster, the hypothesis that would require changing the laws of physics (e.g., rat telepathy or collective consciousness) should be assigned by far the lowest credence, because our knowledge of physics is so secure that even strained mundane explanations (statistical fluctuation, measurement error, exaggeration in retelling) remain vastly more plausible.
“But changing laws of physics to give rats telepathy should be just enormously low on your list of plausible hypotheses for something like this, because we have so much knowledge about how physics works”
The Heisenberg uncertainty principle is not fundamentally a statement about measurement or experimental disturbance; it is a statement about how quantum states relate to observables, namely that complementary observables (like Z-spin and X-spin, or position and momentum) correspond to axes in Hilbert space rotated relative to each other, so no single quantum state can be perfectly localized along both simultaneously.
“So the Heisenberg uncertainty principle is a statement about how quantum states relate to observables. It's actually not a statement about measurement or experimental results at all”
When a gauge symmetry is 'spontaneously broken,' it is not actually broken at all, because a gauge symmetry is merely a redundancy of description (like a coordinate choice); the Higgs field acquiring a nonzero vacuum value hides the symmetry rather than destroying it, and this hiddenness is what allows the W and Z bosons to acquire mass despite gauge invariance naively forbidding it.
“when you spontaneously break a Gauge symmetry, it's not broken at all. You can't break a Gauge symmetry. A Gauge symmetry, when you really dig into what it is, it's just a redundancy of description.”
General relativity deviates from Newtonian gravity (e.g., in Mercury's perihelion precession) fundamentally because in GR the gravitational field carries energy and gravity couples to everything with energy, so the gravitational field is itself a source of gravity (gravitons interact with each other), whereas in Newtonian gravity the field is not a source—hence deviations appear where the field is strongest, as for Mercury closest to the sun.
“The gravitational field in Newtonian gravity is not a source of gravity whereas the gravitational field in general relativity is.”
Although no current scientific theory is 100% correct, a good theory is never 100% wrong either—it captures something true about the universe within a domain of applicability that never disappears; Newtonian gravity is rigorously a limit of general relativity, and 'the table is made of atoms made of electrons, protons and neutrons' will remain true even if those are later found to be composite, resolving the puzzle of why we don't assign low credence to all current theories.
“a good theory, a better theory will be somewhat correct, will capture some true behaviour of the universe... it will have some domain of applicability, but it will never entirely go away”
Energy can be extracted from a spinning black hole via the Penrose process by throwing an object into the hole opposite to its rotation, which kicks the remaining object out faster while slowing the black hole's spin; this is bounded, and along the way Hawking's area theorem holds, meaning the event horizon's area always increases.
“you fly near the black hole, you throw something into the black hole in the direction opposite its rotation, so you slow it down, and in response, you are kicked out faster than you went in, and the black hole slows down a little bit”
The mass-energy radiated as gravitational waves in a black hole merger does not come from inside the black holes (Hawking's area theorem guarantees the merged horizon area, and thus the contained information, is at least the sum of the originals); the energy comes from the orbital motion of the black holes as they speed up falling together, so no interior information escapes.
“Where the information comes from, it's from the energy elsewhere. Like there's energy in the fact that one black hole is orbiting around the other one”
Scale-free behavior—where statistics look the same at every scale, as in the Ising model at its critical point with infinite correlation length—is special and distinct from emergence in general; a classic counterexample is quark confinement, where a definite QCD energy scale separates free quarks/gluons at high energy from confined hadrons at low energy, demonstrating emergent behavior that is not scale-free.
“there are absolutely examples of emergent behaviour, which are not scale-free, scale-free behaviour is somehow special”
Contrary to the endlessly repeated story that Einstein's main objection to quantum theory was its indeterminism ('God does not play dice'), his genuine and sustained concern was the apparent failure of locality due to entanglement, as evidenced by the EPR paper; the dice remark was a quotable aside, while realism and locality were what he truly cared about.
“really what Einstein cared about was locality... that was the whole point of the EPR paper, that you could not find a locally realist theory. He cared about realism more than anything else. Locality, second most”
The selection-effect ('in-between') version of the anthropic principle is trivial in its reasoning but nontrivial in its implications: if physical conditions vary across a multiverse, intelligent beings necessarily find themselves where they can exist, which would explain a small cosmological constant—but because we don't know whether the antecedent (a multiverse) is true, the reasoning, though valid, isn't known to be the correct explanation, so both the anthropic and the unique-predictive-theory approaches remain on the table.
“the anthropic principle... The reasoning in the first example is 100% valid, it makes sense, but we don't know whether the antecedent is true. We don't know if there is a multiverse out there.”
An arrow of time arises automatically and naturally only if a system is both eternal in time and unbounded in configuration space (e.g., a finite number of particles in an infinitely large universe for infinite time); a system eternal but bounded in configuration space merely repeats, so the combination is essential—this is the shared big-picture mechanism behind both Carroll-Chen's model and Barbour's Janus Point, though their specific implementations differ entirely.
“if you have a finite number of particles in an infinitely big universe for an infinite amount of time, then you always get an arrow of time automatically”
The biggest advance that would have informed the thinking of Bohr, Einstein, and Heisenberg is a systematic understanding of decoherence; the Copenhagen philosophy of Bohr and Heisenberg largely shut off investigation into it, and decoherence follows directly from two parts of quantum mechanics present from the start—the single universal wave function and Schrödinger evolution entangling subsystems with their environment.
“I think the answer is pretty obviously a better understanding of decoherence”
The claim that reality is a 'consciousness-driven digital construction' is not simpler than quantum theory; quantum theory is in fact extremely simple (arguably too simple), while invoking consciousness introduces something poorly defined and far more complicated—the appearance of simplicity comes precisely from being ill-defined, whereas a good scientific theory is unbendingly precise, specifying a mathematical structure that makes predictions.
“the idea that right off the bat saying that reality is a consciousness-driven digital construction is somehow simpler, just seems completely wrong to me, completely backwards”
The impression that the universe is remarkably 'tuned' to be discoverable (e.g., via stellar spectra or simple planetary orbits) is ex post facto reasoning analogous to a puddle marveling at how well it fits its hole; we are simply impressed by whatever we can discover first, and in a different universe we would call different discoverable features amazing—there is no a priori standard against which to judge how easy discovery 'should' be.
“We're impressed by the stuff that we can discover first, if the laws of physics or the situation in the universe were very different, there's other stuff we would discover first, and we would call that amazing.”
The zombie argument for non-physical consciousness is self-defeating because if zombies (physically identical beings without conscious experience) are truly conceivable, they would report being conscious and even write books defending consciousness—since reporting is a behavior obeying physics—which means introspection is entirely unreliable; thus any 'consciousness' that has no effect on behavior is not the consciousness we actually mean.
“if zombies are conceivable, then introspection is entirely unreliable, because the zombies are reporting their results of introspection also, and they're completely wrong”
Inflation remains the leading idea about the early universe and the origin of cosmic structure, and gains weak support from a subtle prediction: cosmological perturbations should be almost but not quite scale-free, slightly tilted (larger on larger scales) because the inflaton rolls down its potential producing smaller perturbations at later times—and this tilt is exactly what is observed, though direct evidence has not improved much since the 1998 discovery of dark energy.
“what inflation actually predicts is almost but not quite scale-free perturbations. And that is exactly what we observe.”
Long-term existential calculations of the form 'large number X times tiny probability Y' (e.g., huge harm of extinction times small probability) demand epistemic humility, because the error bars on those numbers are generally large enough to wash out the apparent effect, so such calculations should not drive fundamental decision-making.
“there's a certain amount of epistemic humility that is required when you're calculating things of the form X times Y, where X is a number much, much larger than one, and Y is a number much much less than one”
Top research institutions select faculty almost entirely on research brilliance and accomplishment, caring essentially nothing about whether candidates are kind, generous, or good teachers; therefore any correlation between academic success and unkindness is better explained as a filtering effect than as success grinding people down, and academic success should if anything make generosity easier.
“academia does not select on kindness or generosity. When universities hire people to be on their faculty, they do not say, 'Oh, we want people who are kind and generous.'”
Not enough effort goes into seriously considering that the Big Bang was neither the beginning of time nor a special event; many proposed pre-Big-Bang scenarios still attach specialness to the Big Bang (mirror-image contracting universes), but it could instead have been quiescent empty spacetime that bubbled off a universe, or something radically different.
“I think that not enough brain power is put on, considering the possibility that the Big Bang was not the beginning and it wasn't a special event either.”
Naive utilitarianism is flawed because you can always win the moral calculus by adding more people: a quadrillion miserable lives can be made to 'outscore' a short flourishing civilization, which is unacceptable; moreover, the impossibility of defining a well-defined measure on infinitely large sets means counting all future people equally would imply the only moral act is preventing the heat death of the universe, which is absurd.
“That's the fundamental problem with utilitarianism; you can always win just by adding more people to the game, and I think that's wrong.”
Emergence works by coarse-graining: grouping the vast microscopic degrees of freedom of a system into a small number of observable features (like temperature, pressure, velocity in a fluid) that contain far less information yet still yield a highly predictive theory of how the system evolves.
“That's what emergence is all about, the ability to coarse-grain, that is to say to only take into consideration some very tiny subset of all the data, and nevertheless get a theory that tells you something useful about how the system evolves over time.”
The common claim that special relativity cannot deal with accelerating particles or frames is false; special relativity has no problem describing accelerated particles, and what is actually meant is only that there is a preferred class of inertial reference frames in which the physics looks simpler—you can still use whatever coordinate system you want.
“you will often read the completely incorrect claim that special relativity can't deal with accelerating particles... That's completely BS.”
Imaginary time is just a convenient mathematical trick with no deeper truth about the nature of time itself; it originates not in general relativity but in the WKB approximation in quantum mechanics from early explorations of quantum tunnelling, and one should not be fooled into reifying it when translating it into words.
“No, it's just a convenient mathematical trick.”
Alternative ('tired light') explanations for the cosmological redshift fail empirical tests because the relativistic expansion explanation makes distinctive predictions—notably time dilation of supernova light curves proportional to redshift, which is observed, and a perfect black-body microwave background spectrum that scattering off matter would distort—both of which match observation.
“the redshift because of the expansion of space comes along with a time dilation... that makes a very strong prediction for the shape of the light curve, of that supernova, it should also be time-dilated, guess what? It is”
The continuity of personal identity over time is not an illusion but a real, useful, higher-level emergent description—just like the continuity of a chair—because grouping the world into persistent entities provides a compact, predictive handle on reality; an illusion is when X is believed true but is false, whereas emergent identity is both useful and true.
“that's the sense in which personal identity is not an illusion, it's real, it's continuous over time in that sense”
In emotionally charged debates, recognizing the issue of losing one's cool is itself most of the solution, because being reflective and self-aware lets you correct course over time; combined with correcting for innate tribalism (the tendency to frame discussions as competitions to win rather than mutual learning) and disengaging from people who can't be reasoned with, trying harder already places one among the best interlocutors.
“the fact that you recognise the issue is 98% of the solution to the issue”
Quantum computing demonstrably works—small-qubit algorithms have been run on real machines—though current small devices are slower than classical simulations; it is extremely unlikely that some modification of quantum mechanics prevents quantum computers from achieving speed-ups at scale, and anyone predicting otherwise should make that prediction precise since confirming it would overturn quantum mechanics and earn a Nobel Prize.
“It's certainly not possible the quantum computing doesn't work at all because it works. It's been done”
Black holes can be distinguished from other massive objects like neutron stars by simultaneously measuring mass and size (if smaller than its Schwarzschild radius and GR is roughly right, it must be a black hole) and by analyzing emitted light—a non-black-hole central object would glow under heating from its accretion disk—though 100% certainty is impossible without direct inspection.
“if the size of the thing is smaller or as small as it's short shield radius, there's literally nothing else it can be other than a black hole, if general relativity is roughly right”
Values like honour and valour are generally good in themselves; the real issue is that society is changing rapidly relative to past eras, so people taught older value systems find themselves mismatched with the world they inhabit, and the productive response is to reason with and explain to such people rather than simply condemn them—while remaining alert that some advocates use these positively-valenced words to advance unrelated agendas.
“because society is changing in ways that it's a pretty rapid change in how we live our lives compared to previous years... people are left behind a little bit, or people still taught old ways of thinking or valuing, even though the world is changing around them”
Contrary to the mainstream framing, the real puzzle is not how to accept non-locality but why the universe looks even a little bit local: generic Hamiltonians are non-local, and gravitational effects combined with quantum mechanics naturally produce non-locality, so non-locality is perfectly natural and locality is the surprising feature requiring explanation.
“I think the puzzle is why there is locality at all, not the puzzle of why there's non-locality at all”
To the extent God is invoked as an explanation for features of the universe, God can and should be evaluated like a scientific theory—on simplicity, fit to data, and alternatives—and it fails: it commits the cheat of exempting God from the 'everything needs a cause' principle, and the features the universe actually has differ from what one would predict if God were responsible.
“to the extent that you attribute to God explanatory powers, you can also evaluate God as a scientific theory”
As a general pattern, when human beings gain the ability to do something, they generally do it—so debating whether we should de-extinct species is less consequential than preparing to deal with the fact that it will happen in some cases.
“when we gain as human beings the ability to do things, we generally do them”
There is no intrinsic virtue or vice to de-extinction because ecosystems should be understood as dynamic—species will always go extinct and new ones arise—so neither bringing species back nor preserving every species can be the goal; given the law of unintended consequences in complex systems, de-extinction should be approached with extraordinary caution.
“I think that the environment, the ecosystem, all of the different diversity of life that we have has to be thought of as dynamic”
Although supermassive black holes are today a tiny fraction of a galaxy's mass, over sufficiently long timescales most matter in a galaxy will fall into its central black hole before the black hole evaporates via Hawking radiation, while some stars are ejected by slingshot encounters—so the far future is plausibly dominated by sleepy black holes plus some wandering ejected objects.
“if you just wait a very, very long time, eventually all that stuff is gonna fall into the black hole, and the question is, 'Which happens first?'”
The high-level, non-cynical purpose of an institution like Harvard is to produce good research and scholarship, not to educate students; students exist to enable that purpose, and if Harvard switched to prioritizing teaching it would cease to be Harvard—neither good nor bad, just different.
“the non-cynical purpose of Harvard is to do good research, to do good academic scholarship and to produce great things, that's the purpose. Students are there to sort of allow that purpose to happen”
Given the long history of unfulfilled promises that fusion power is imminent, neither extreme is justified: one should not believe new fusion timelines merely on trust, nor conclude fusion will never arrive merely because it has always been promised; credible optimism requires an explicit understanding of why past predictions were wrong and why those reasons no longer apply.
“It would be much more convincing if you could actually understand why they were wrong the first time and why those reasons do not apply this time.”
Pro-free-speech arguments fail when made purely procedural—defending someone's right to speak without disclosing the substance of what they said; since even the freest platforms have substantive restrictions, the genuine question is where to draw lines and who decides, and refusing to confront the content of speech is a form of cheating.
“they make an entirely procedural. They completely ignore the substance of what is going on.”
There is no a priori impossibility in deriving first-person experience from physics: under poetic naturalism, 'I am experiencing the redness of red' is a higher-level emergent way of talking about a collection of particles and fields behaving in certain ways, and one can readily imagine particles and fields that would exhibit all the behaviors associated with consciousness, so the hard problem will not be solved but will evaporate once we map experiences to brain states.
“I have zero problem imagining that a bunch of particles and fields would do that, I don't see what the obstacle is”
Pilot-wave (Bohmian) interpretations are Carroll's second-favorite but objectionable because their hidden variables and the spontaneous-collapse models alike get their plausibility by smuggling in pre-existing classical notions—particle positions and locations in space—none of which will survive once we understand quantum gravity and emergent spacetime; Everettian quantum mechanics avoids this by accepting only the wave function.
“both the hidden variable models and the spontaneous collapse models get a lot of their mileage out of smuggling in pre-existing classical motions of stuff and locations and space and all these things, none of which are gonna survive once we understand quantum gravity”
Gödel's incompleteness theorem is irrelevant to scientific practice because Gödel's theorem concerns what can be proven within a formal system, whereas science does not prove things—it proposes hypotheses, tests them, and updates credences based on fit, simplicity, and fruitfulness—so the relevant escape is not extending axioms but recognizing that proof is not the scientific aim.
“Godel's theorem is a statement about what can be proven inside some formal system, and science is not about proving things, so who cares?”
Psychoactive substances like LSD do not fire up the brain to produce new visions; they lower the brain's inhibitions or filters—like removing a noise filter—revealing what is already going on in the brain, and Carroll found LSD of no help for serious theoretical physics since one cannot even write down equations on it.
“the real trick of these substances is not that they fire up your brain and give you all these visions, is they lower the inhibitions in your brain, they sort of lower the filter”
The scenario that life may be ubiquitous but rarely highly technological or conscious is, in some sense, the simplest explanation for the Fermi Paradox—we haven't found intelligent life because intelligent life is extremely rare, either because life itself is rare or because intelligence is rare among life—though the claim that destroying intelligent life on Earth would eliminate meaning in the galaxy, while sympathetic, lacks a clear intellectual argument.
“it's, in some sense, obviously the simplest explanation for the Fermi Paradox. Why haven't we found intelligent life elsewhere in the universe? Well, because it's not there”
Because social media platforms uniquely pretend to give everyone a voice, are few in number, and exhibit winner-takes-all dynamics that prevent painless switching, the newspaper analogy only goes so far, and one could imagine treating them like regulated utilities with written-down laws specifying which speech is protected and which the platform may exclude.
“I could imagine an argument that you should start treating social media kind of like utilities, kind of like the phone company or whatever, and have some written-down laws that tell us what kinds of speech are absolutely protected on these platforms”
Many-worlds does not posit many worlds as an additional assumption; rather, taking the wave function and the Schrödinger equation seriously and literally, decoherence causes subsystems to become entangled with their environment, splitting one world into multiple branches—so the worlds are derived, not assumed, and the measurement problem is resolved by ordinary Schrödinger evolution without consciousness or collapse.
“What's important is that the Schrödinger equation tells you how the wave function evolves, and it evolves from having only one world in it to having multiple worlds in it. That's how you resolve the measurement problem in many worlds.”
In Carroll's personal observation, non-religious family members handled the death of someone close better than religious ones, suggesting that genuine acceptance—fully accepting that the person is gone rather than maintaining a 'halfway house' belief in continued existence one cannot access—aids grieving, though he stresses this is anecdotal and may not generalize.
“the people who were not religious uniformly handled it much better... it spoke to the fact that acceptance matters”
Questions like 'why am I me rather than Isaac Newton' are grammatically correct but make no sense and dissolve under investigation; there is no other person you could be, so the question is trivially empty—similar to claiming to be a reincarnation with all memories wiped, which conveys nothing.
“I think that the question, 'Why am I me?' rather than 'Why am I Sir Isaac Newton?' makes no sense whatsoever, not even a little bit of sense.”
Personal-identity puzzles arising from duplication, teleportation, or copying experiments cause trouble only because such situations never occur in the real world; our concepts of identity were developed for ordinary experience, so we should not demand they apply consistently to unprecedented circumstances—and for the robot-copy case there is simply a fact of the matter that there is a robot in room one and a robot in room two, with the copied robot merely fooled.
“when you violate the things that actually we're familiar with in the real world, you come across situations which don't fit comfortably with our categories”
Reasoning based on impossible hypotheticals (like 'if we were Laplace's demon' or 'the length of a space-like trajectory') is not invalid; such counterfactuals are colorful, legitimate devices for illuminating concepts—Laplace's demon is just a vivid metaphor for determinism—and must be understood in the spirit in which they are offered.
“No, I don't agree, because you have to take the spirit... You have to make a statement like that in the spirit in which it is offered”
Roughly speaking, you cannot build a thermodynamic engine that does useful work from cosmological expansion, because the energy of expansion is similar to objects simply moving apart from each other—you could extract a finite amount of energy (as with two objects connected by a spring moving apart), but it is not a limitless supply and is impractical at cosmological scales.
“The cosmological expansion is similar in many ways to what you would have if spacetime itself were not expanding, but matter was just moving apart as if from an explosion”
Many features arising through biological evolution are not necessarily adaptive (spandrels): a trait may exist not because it confers a reproductive advantage but because it accompanies another advantageous feature or shares genes with one—so it's unclear whether laughter has a reason for existing or is merely a byproduct of physiological features adapted for other reasons.
“spandrels being this idea of Stephen Jay Gould, Niles Eldridge and thinking about evolution, that many things that happen according to biological evolution are not necessarily adaptive”
There is no moral superiority in taking public transportation; rather, cities and communities should design public transit so that it is the easier, more convenient option, making people want to take it—Carroll prefers trains only when the system is organized to make them genuinely convenient, not out of duty.
“I'm someone who thinks that cities and communities should design public transportation so that people want to take it, so that it is the easier, more convenient thing to do”
Finding parts of a subject to be a chore is not a sign to abandon it; mastering any field requires tedious practice (manipulating tensors, shooting free throws, playing scales, memorizing vocabulary) as an unavoidable part of a larger achievement—'there is no royal road to geometry'—though if an entire field feels mostly like a chore, switching fields is a separate, more difficult question.
“we have to sometimes grind, we have to sometimes do chores”
In theoretical physics, when a proposed model fails it usually goes unpublished, unlike a null experimental result which is informative to others; the time to recognize a dead end is highly inconsistent, ranging from a day to over a year, and ideas can have long gestation, becoming publishable years later when technique or motivation catches up.
“for theorists like myself, most of the time when we come up with an idea and it does not pan out, it just goes unpublished”
There is no reliable algorithm or conscious strategy for making new friends as an adult; friendships form unpredictably through many heterogeneous channels (work, online, random events), so the only tangible advice is to put yourself out there, accept some embarrassment risk, and go through many people before finding ones you genuinely connect with.
“I don't think that there is sort of a strategy that you consciously adopt to make new friends”
Greatest-of-all-time debates across sporting eras are fun but meaningless because athletes from different eras trained differently and played different games (e.g., no three-point line in Chamberlain's day) making direct comparison impossible, and because what ultimately matters in sports is who won the actual game played that year, not hypothetical superiority.
“you can't compare people across eras... They live in different eras. They were trained in different ways, they played different games, there was no three-point line Wilt Chamberlain played”