Astral Codex Ten

cancer

46 min read

cancer

Article

cancer is a recurring concept in the Astral Codex Ten archive, appearing 4 times across 4 issues between February 16, 2021 and September 04, 2025. The archive places it in contexts such as “where sunlight matters a lot but study after study has shown Vitamin D doesn’t help at all”; “we put a lot of time and money into researching cures for cancer”; ““is cancer a disease?” is a political question”. It most often appears alongside Australia, NYT, US.

Metadata

  • Category: Concepts
  • Mention count: 4
  • Issue count: 4
  • First seen: February 16, 2021
  • Last seen: September 04, 2025

Appears In

Source Context

Recovered passages from the original issue text. When the raw archive preserved outbound links inside the source passage, they are listed directly under the quote.

Vitamin D: Much More Than You Wanted To Know
February 16, 2021 · Original source
The loose ends that bother me the most are the seasonal pattern, the latitude data, plus the increased risk of hospitalization and death in Asians. I don't have a great explanation for those. One possibility is that sunlight does help prevent coronavirus, it just isn't Vitamin D mediated. I suspect that the "anything involving sunlight is Vitamin D" assumption a lot of epidemiologists have isn't going to hold up very well - this seems especially true for cancer, where sunlight matters a lot but study after study has shown Vitamin D doesn't help at all. It may also be true for schizophrenia, although I'm going off one really pathetic study there and it could very well turn out to be Vitamin D after all. Some people are doing a little bit of work to clear up what the sunlight-related-Vitamin-D-independent pathways might be; I think nitric oxide has come up a few times.
Inline links: nitric oxide
Book Review: Lifespan
December 02, 2021 · Original source
David Sinclair - Harvard professor, celebrity biologist, and author of Lifespan - thinks solving aging will be easy. “Aging is going to be remarkably easy to tackle. Easier than cancer” are his exact words, which is maybe less encouraging than he thinks.
Inline links: Lifespan
What if you’re not a mouse, and you live in one of the 100% of countries that have banned random irresponsible genetic engineering on humans? Sinclair thinks there’s still hope. Sirtuin activity seems to be regulated by a protein called mTOR (motto: “The Protein That Regulates Everything”, see discussion of its role in depression here, obesity here, cancer here, etc). In times of plenty, mTOR switches on, causing cells to divide and grow. In times of deprivation, mTOR switches off, causing cells to “hunker down” and go into power-saver mode. Apparently part of power-saver mode is damage control, so this turns on the sirtuins and makes them do more epigenetic repair, keeping you young. All you have to do to stay younger longer is keep mTOR from activating.
Inline links: here, here, here
That doesn’t mean we can’t be immortal, though. Sinclair’s lab has another research program, focusing on stem cells. These produce new, epigenetically healthy cells wherever they’re placed. And we now know that with a couple of chemicals called Yamanaka factors, you can make adult cells become stem cells again. Sinclair’s idea is to genetically engineer triggerable Yamanaka factors into the cells of human adults. Then, when you’re starting to feel old, you trigger the factors, some of your cells revert to stem cells, and they replace your old decaying cells with epigenetically healthy ones. Every biologist I mention this to has the same objection, which is “won’t that make you have every kind of cancer at once?”, and, in their defense, the first hundred times Sinclair tried this his mice got some pretty crazy cancers. But he swears they have solved this problem and the mice are doing fine now. Some of them are living about 40% longer than normal, which I notice still isn’t immortal but seems like a step in the right direction.
Sure, Whatever, Let's Try Another Contra Caplan On Mental Illness
June 29, 2023 · Original source
Left: my position. Right: my position, “rounded off” to Caplan'’s position In particular, he claims I am FORCED to either accept that all mental illnesses are just “preferences” and so not illnesses at all, or as posited in a response by Emil Kierkegaard, that homosexuality is a mental illness and therefore bad. You will not be surprised to learn that I don’t think of myself as secretly admitting this, or forced into doing anything. II. Bryan mentions how I have already addressed his fork with a much more in-detail discussion of how we classify something as a disease or not at this link, to which I would add this post as fleshing out the same framework. Put simply, declaring something a “disease” is a complex category-boundary-drawing issue that combines facts and values, just like all category-boundary-drawing issues. I said that it’s a political question whether or not you classify homosexuality as an illness. Caplan thinks of this as some sort of incredibly deep concession. But it’s a political question whether or not to classify any condition, including physical conditions, as illnesses. It’s just that the political question is usually very easy. This shouldn’t be surprising - most political questions are easy! “Should we set every tree in the United States on fire, then dump the entire Strategic Uranium Reserve in the Mississippi River?” - this is a political question, in the sense that you could propose it for a vote and people would have to form an opinion on it. It doesn’t show up on C-SPAN because it doesn’t satisfy anybody’s values. It’s a political fight where one side has a constituency of zero. In the same way, “is cancer a disease?” is a political question. Maybe cancer makes you cough up blood and die. Basically everyone is against this, so it’s easy to condemn it and agree that doing it is worse than not doing it. If for some reason there were some strong political constituency in favor of coughing up blood and dying, who thought were were unfairly stigmatizing this wonderful prosocial activity, then we would have to have a political fight about it. This fight would have to involve comparing values (eg being against death) rather than comparing facts (eg cancer is caused by a mutation in such and such a gene).1 (see also: The Tails Coming Apart As Metaphor For Life and Ambijectivity. Categories often contain a simple region where they operate perfectly and where it would be perverse to consider them a political question even though they sort of are, and a more complex region where they start to break down and we have to agree on some final border) Is Down Syndrome a disease? It often causes poor health and low IQ; I’m pretty against both of these things, so I would say yes. Still, there are a bunch of people who argue it isn’t, maybe because they don’t care what your health or IQ is, or because they think stigmatizes Down Syndrome patients. I think these people are wrong, but only in the same way that I think people who support the Russian invasion of Ukraine, or who hate free speech, are wrong: they have bad values, they’re against human flourishing, they’re on the wrong side of a political question. Is depression an illness? It causes you to be miserable and not be able to do most of the things you want to do. Same story. I can’t imagine anyone being in favor of this, and I hope there’s a broad base of support to continue classifying it as an illness - but it’s a value judgment. Caplan says okay, maybe sometimes in some ways the category boundary drawing is hard, but he proposes a bright-line rule: No preference is a disease. No matter how bizarre or horrible (or common or wonderful). Diseases are constraints, not preferences. Part of my frustration with Caplan is that I feel like I have proven this constraint/preference distinction incoherent and misleading again and again over the course of our “debate” and he’s never responded. He just keeps saying “but the constraint/preference distinction!” For the sake of completeness, I’ll give my summary of what he thinks the distinction is, plus four of what I consider to be the strongest counterarguments. My interpretation of Bryan’s theory (I’m putting this in a quote block to specify I’m devil’s-advocating it, but this is my summary and not his): If we think like behaviorists, all we can really see about mental illnesses are unusual behaviors. For example, a depressed person stays in bed all day and doesn’t work. An alcoholic drinks himself to death. A psychotic person runs out in the street naked claiming to be God. These seem like choices. You can imagine the depressed person choosing to throw parties and work hard instead. You can imagine the alcoholic choosing to throw out his beer and never drinking again. You can imagine the psychotic person choosing to put on his clothes and act normally. In fact, if you put a gun to the alcoholic’s head and threatened to shoot him if he ever drank again, probably he would stop drinking. Therefore, we should model these conditions as unusual preferences/choices, not as diseases. The hallmark of a disease is a constraint, something you cannot “choose” to overcome, something you couldn’t overcome even with a gun to your head. For example, a paralyzed person cannot choose to walk no matter how hard she wants to, or how dire the consequences for not walking. Therefore, paralysis is an unusual constraint, and depression is an unusual preference. We may choose (for political reasons) to stigmatize certain unusual preferences. Maybe the people who have them will choose (for signaling reasons) to cooperate in their own stigmatization. But realistically these are just completely voluntary preferences. If we don’t like them, we should ask the people who have them to choose differently, instead of treating them as diseased. My counterarguments: — 1: Counterargument From Physical Illness, Part I The simple preference/constraint model clearly doesn’t describe mental illness very well. But it’s actually much worse than that. It doesn’t even describe physical illness. Consider a migraine. If we think like behaviorists, all we can really say about migraines is that someone locks themselves in a dark room, clutches their head, and says “oww oww oww” a lot. If we put a gun to a migraneur’s head and threatened to kill them if they didn’t go to a loud party, they would grudgingly go to the party. So clearly (says a hypothetical version of Caplan, whose answers I must rely on because the real Caplan has never addressed this objection) migraine headaches are a preference, not a disease! Some people just like locking themselves in dark rooms, clutching their head, and saying “oww oww oww” a lot! If other people call this a “disorder”, they’re choosing to stigmatize migraineurs; if migraineurs agree it’s a disorder, they’re just trying to escape responsibility for their antisocial choices. You could say the same about many - maybe most - physical diseases. Why not say that chronic pain is just a preference for grimacing a lot? That itchy rashes are just a preference for scratching yourself a lot? That colds are just a preference for lying in bed and blowing your nose a lot? (I believe most people with colds could get up, go to work, and avoid blowing their noses, if their lives depended on it). Or we could stop thinking like behaviorists, a philosophy which nobody has taken seriously since the 1970s. Once we agree that people are allowed to have internal states, and that the rest of us are allowed to acknowledge those internal states, the paradox disappears. We can agree that the essence of migraine headaches is pain, especially pain in response to strong sensations. The essence of itchy rashes is a feeling of itchiness, which is relieved when we scratch it. The essence of colds is feeling unwell and ugh and wanting to stay in bed and having unpleasant congestion in your nasal passages. None of these particularly change your preferences. Both I (never had a migraine) and the average migraineur have a preference for not having our head be in terrible pain. But the migraineur needs to avoid bright lights in order to satisfy this preference, and I don’t. So she very reasonably avoids bright lights. Once we’ve admitted this, it’s natural to also admit that depression involves negative emotions and low energy, that alcoholism involves a craving to drink alcohol, and that psychosis involves disturbed reasoning processes which make running out in the street naked claiming to be God seem like a good idea (all with other preferences intact). This is more parsimonious than Caplan’s theory, better matches the testimony of the mentally and physically ill themselves, and doesn’t require the mentally ill to be running some 4D-chess-style network of lies (such that actually the psychotic person’s reasoning is completely normal and they’ve just managed to perfectly trick everyone into thinking that it isn’t and tell a perfectly consistent story all the time and stick with their deception even when it presents an extreme threat to their life and freedom). — 2: Counterargument From Gradients Preferences and constraints naturally shade into each other. Let me give three examples. Example 1: I am a mediocre runner, able to run about 5 km before getting tired and stopping. One day, at exactly the 5 km mark, a demon appears before me, and says it will kill me unless I run another 1 km. I’m pretty upset by this, but I gather all my willpower, try really hard, and manage to run another 1 km. Then the demon appears again and says haha, I was just joking last time, but now I’ll really kill you if you don’t run another 1 km. For some reason I’m gullible, I believe it, and even though I am in extreme pain I make a herculean effort and run another 1 km. Again the demon appears and makes the same threat, and this time I say sorry, I really can’t run another inch, guess I’ll die. The demon says okay, new threat, it will kill me and my entire family horribly if I don’t run another 0.1 km, but give me $1 million if I do. I call upon some kind of reserve of courage worthy of the heroes of old, put one foot in front of the other, and make it a final 0.1 km before stopping. Again, the demon says haha, fooled you, you need to run another 0.1 km. I try this, collapse, and await my impending death. Do we argue that I had a simple preference again running 6, 7, and 7.1 km, but that my inability to run 7.2 km was a true constraint? It seems obvious that my difficulty running 7.1 km is of the same type as my difficulty running 7.2 km, and it just passed some threshold where I couldn’t do it anymore no matter how much it mattered. Example 2: The demon puts a dimmer switch on my leg nerves. When it’s at 100%, I have totally normal movement. When it’s at 0%, I’m paralyzed from the waist down. At 25%, I can sort of kind of walk in extreme pain. The demon threatens to kill me unless I succeed, so I shamble a short distance. Then the demon turns the switch down to 24% and threatens me again; I try my best, but fail. I think Caplan would have to say that at every level up to 25%, I simply have a preference against walking, which is fine and voluntary and my own fault and not a disease in any way. Then at 24%, it suddenly becomes a constraint inflicted on me by an outside agency and which I deserve sympathy for. Instead, I would rather describe things that make an action difficult and unpleasant as in some sense real constraints. When the dimmer switch is at 25%, I have an external constraint making walking difficult and unpleasant, although I can overcome this and do it anyway with a strong enough incentive. When the switch is at 24%, it’s become so difficult that no incentive can make me do it. There’s no qualitative boundary, just a quantitative one. Example 3: Try to hold your breath as long as you can (please don’t go overboard and hold it so long you pass out). If your experience is like mine, at each moment you’ll feel like - given a slight exercise of willpower - you could choose to hold your breath one more second if you so desired. But if your experience is like mine, you will also find that no amount of love or money could make you hold your breath successfully for (let’s say) three minutes.2 Is there a point where not wanting to hold your breath any longer switches from a preference to a constraint? Or have you discovered a place, in the dark moments just before suffocation, where these concepts lose all meaning? — 3: Counterargument From Physical Illness, Part II Caplan claims that mental illnesses involve preferences and physical illnesses involve constraints. But a second’s thought reveals this is not actually true, even if you accept the whole preference-constraint dichotomy Consider cancer. Cancer involves some constraints; for example, it might kill you, and you cannot choose to live instead, even if someone put a gun to your head and demanded it3. But until that happens, it mostly looks like preferences. People with cancer might stay in bed, saying they feel too sick and weak to get up and do things. But if you threatened them with a gun, they could probably get up and do things. People with cancer might refuse to eat, saying they feel too nauseous and have no appetite. But if you threatened them with a gun, they could probably get down some food. Meanwhile, plenty of mental illnesses include constraints. One of the diagnostic criteria for depression is cognitive and memory problems; people with these problems cannot choose to remember things better, even with a gun to their head. Many people with psychosis cannot speak or reason normally, even if you put a gun to their head and ask them how a healthy person would answer a question. People having panic attacks cannot choose to have a normal heartbeat, or to stop shaking or sweating. Depression and anxiety are both associated with insomnia; try to will yourself to sleep and you’ll sleep less, not more. Both physical and mental illnesses are complex bundles of preferences and constraints, which shouldn’t be surprising given that preference vs. constraint is an oversimplified distinction that breaks down outside its legitimate domain. — 4: Counter-Argument From The Gun-To-The-Head Test Actually Not Working A depressed person may not be able to get out of bed or live a normal life. This might get so bad that they decide to commit suicide by shooting themselves in the head. Confronted with a choice between living a normal life, or a gunshot to the head, they have chosen the gunshot4. It appears that they have passed the gun-to-the-head test that Caplan loves so much. I feel bad including this one, because Caplan can fairly object that this is just another preference. Maybe depressed people completely voluntarily choose to lie in bed for a few years while falsely claiming to be miserable and then shoot themselves in the head, and all of this is a perfectly free choice that they are happy with. I cannot disprove this, only point out how unparsimonious it is. Maybe a better example is when a psychotic person attacks the cops, the cops order him to stop or else they’ll shoot him, the psychotic person continues attacking them (eg because he believes he’s invincible) and then the police go ahead and shoot him. Again, Caplan could say that this is just a preference for attacking cops and then being killed. But in that case he should stop touting the “gun to the head test” as meaningful. Rather, he should admit that his theory is completely unfalsifiable - no matter what actions a mentally ill person does, what tests they pass or fail, he can just say they had a preference for doing whatever they did. In fact, at this point I don’t see why he even has to acknowledge the existence of constraints at all. One might as well claim that a paralyzed person could walk if they wanted, but chooses not to. III. I think Caplan is groping towards something like the following criticism: Suppose we simplify depression to “person lies in bed and doesn’t do anything all day”. Caplan’s model treats this as “depressed person has preference to lie in bed”. My model treats this as “depressed person has an abnormal mental/emotional/motivational state that makes it difficult and unpleasant for them to not lie in bed”. Now we consider a gay person. Caplan’s model treats this as “person has a preference to be gay”. Wouldn’t my model have to treat this as . . . person has abnormal mental/emotional/motivational state that makes it difficult and unpleasant for them to be heterosexual? In some sense this is true. We could imagine some very religious man from the 1950s who really wants to be straight, marry a woman, and raise a family. But due to some hormonal disturbance, he feels a very strong urge to have sex with men. How is this different from (let’s say) depression-secondary-to-hypothyroidism, where some person really wants to live a normal life, but instead, due to a hormonal disturbance, feels unable to do anything but lie in bed? It doesn’t seem that different to me. It also doesn’t seem that different from a straight guy who wishes he were gay (maybe for LGBTQ cred, or because it would make it much easier to find partners) but feels a very strong urge to have sex with women. So does that mean that depression is “just a preference”? I don’t think so, because none of these scenarios seem that different from the person with the migraine either! I think the preference/constraint dichotomy is a bad way to think about about this whole class of things. I think all of the following things shade into each other: A migraine. You could think of this as a preference for sitting in a dark room and saying “ow ow ow” - or as an internal state of head pain.
Inline links: as posited in a response by Emil Kierkegaard, this link, this post, 1, The Tails Coming Apart As Metaphor For Life, Ambijectivity, 2, 3, you’ll sleep, 4
Isn’t cancer obviously, nonpolitically, an illness, in the sense that it clusters with all the other things we agree are illnesses? I think there are a few answers to this. One is that I would say Down Syndrome should also be in this category, but there are some real people who claim it’s not a disease - just because I would *like for* something not to be a political question, doesn’t mean it isn’t! An even better example is aging - which involves just as much biological dysfunction as cancer, is even more fatal, and yet - because some people aren’t against it and think death is good in this one weird case - it’s hard to get it classified as a disease. Most sane classifiers, presented with all diseases other than cancer and asked to determine whether cancer is inside or outside of the category, would say inside - but it’s our job to enforce that in real life, and neither “sane classifier” nor “agreeing on all other diseases” are trivial.
Inline links: it’s hard to get it classified as a disease
Or can you? There are lots of stories of some cancer patient who “holds on” until their beloved spouse can be by their bedside. Is death, like taking a breath, something you can put off voluntarily, but whose delay requires more and more willpower until everyone eventually gives in?
Links For September 2025
September 04, 2025 · Original source
Note: percentages are of total, not of each row! 29: Related: social science team proposes a three-stage model of secularization: decreased public ritual participation → decreased personal importance → decreased identification, presents apparently confirmatory data. If true, would be somewhat inconsistent with intellectual models (eg people learn about evolution and start doubting the Bible) and more consistent with institutional models (eg the government provides welfare so people no longer need to be part of a tight-knit church). 30: Navigating LLMs’ spiky intelligence profile is a constant source of delight; in any given area, it seems like almost a random draw whether they will be completely transformative or totally useless. Now Ethan Strauss reports that they are, for some reason, extraordinarily effective at teaching people golf. “I am predicting the Golf Revolution, or perhaps decline, if your perspective is that optimization tends to ruin hobbies. A sport for obsessives has been gifted the ideal tool for refinement.” 31: Claim (via nxthompson on X): “In a huge survey of young kids about phones and technology, they all say they want to be out playing in the real world. But parents don't let them out unsupervised. So they're stuck on their phones.” Interesting, but I’m nervous about social desirability bias - how many adults would say on a survey that they would rather be on their phones than playing with friends? But adults do have this choice and mostly go with the phones. 32: Steven Adler on AI psychosis. He tries to analyze ER admissions data for psychosis and finds no change. I don’t think anyone reasonable expected this to be a large enough effect to show up in ER admissions data, but there are lots of unreasonable people so I appreciate his effort. He thinks AI companies might have better data on this, and encourages them to release it. 33: Cuartetera was the greatest polo horse ever. Polo players responded in a very practical way: they cloned her, dozens of times (and it worked; the clones are also excellent). Now there is a lawsuit as different polo teams fight to get their hands on Cuartetera clones. What is the equilibrium? If the outsiders get their hands on the genetic material, do we see a world where every polo horse is a Cuartetera clone? How much is lost if nobody ever tries to breed a polo horse better than Cuartetera (since the economics might not check out if the odds of success for any given foal is too low)? H/T Gwern and Siberian Fox (on X). 34: Claim: as of 2013, India’s Agarwal caste, who make up less than 1% of the population, got 40% of the e-commerce funding. 35: Owlposting: What Happened To Pathology AI Companies? Pathology is a medical specialty. A typical task involves looking at a microscope slide full of cells and trying to determine if any of them are cancerous. This seems like a good match for AI - and for years, studies have been showing that in fact AI can equal human experts. So why isn’t it being used more? The author’s three answers: first, slide scanning is expensive and clunky, and you can’t apply AI to a slide until you digitize it. Second, it’s hard to figure out a business plan where this saves someone money and doesn’t step on the toes of big companies that can outcompete anyone they don’t like. Third, pathologists use the context of a patient’s entire clinical history when they interpret a slide, and AIs that can’t do that (either because of technical limitations or legal/privacy limitations) are at a disadvantage even if their skills specifically relating to slide-reading are better. 36: Noahpinion: Will Data Centers Crash The Economy? Suppose that AI is a bubble, either permanently (because the technology isn’t really transformative) or temporarily (because it can’t transform things quickly enough to keep up with all the dumb money pouring into it). Will the sudden write-off of data centers lead to a broader economic collapse? In 2001, the dot-com bubble harmed the tech sector, but didn’t take the rest of the economy down with it; in 2008, the subprime mortgage bubble did take the rest of the economy down with it, because it damaged banks that the whole economy relied on. The optimistic case for AI is that data center spending is mostly coming from big companies like Google and Meta that can absorb a lot of loss. The pessimistic case is that some of the money is coming from private credit, a new-ish form of finance which hasn’t really been stress-tested and whose failure modes are still poorly understood. Noah’s final verdict: the stage isn’t obviously set for a crisis yet, but there’s the potential to get there and we should consider acting (how?) early. 37: The latest Twitter talking point is that universal hepatitis B vaccination at birth is “woke”: Hep B is (aside from mother-to-child transmission) often sexually transmitted, slutty women’s children are more likely to have Hep B, so perhaps giving the vaccine to everyone (instead of testing and only giving to the children of women who test positive) is an attempt to spare slutty women the embarrassment of getting a positive test. Ruxandra Teslo provides the counterargument - Hep B tests take a while, the medical system is fragmented, and any attempt to test people and then give the vaccine inevitably leads to many positive tests falling through the cracks. Vaccinating at birth is easy and hard to screw up, the vaccine has no known side effects, and empirically child Hepatitis B rates go down (by as much as 2/3!) when countries switch from test-and-vaccinate to universal vaccination. This benefits everyone - even people who never have unprotected sex and always follow up on their medical tests - because toddlers in daycare exchange saliva copiously, and if your toddler exchanges saliva with a Hep B positive toddler they could get the disease. A funny Twitter interaction was seeing Republicans in Congress hop on the anti-slut anti-vaccination bandwagon - except for Senator Bill Cassidy (R-Louisiana), who happens to be a liver doctor, and who is still fighting the good fight. I am always nervous when a good person who I like starts engaging on Twitter, since it elevates the discourse there but also gradually turns their brain into mush - but Ruxandra has made the leap and is doing a great job not just on bio related topics but also (for example) countering Curtis Yarvin on the history of her native Romania. 38: The response to GPT-5 was confusing; most specific people who reviewed it said they were impressed (Ethan Mollick, Tyler Cowen, Nabeel Qureshi, Taelin), it performed as expected on formal benchmarks, but the overall vibes declared it a big failure. Peter Wildeford speculated that maybe there was some kind of sinister pay-to-play early access bias involved. Zvi went the other way, calling it a “reverse DeepSeek moment” (insofar as DeepSeek was a pretty average model that got glowing praise.) In the end, I agree with Peter that this was mostly a branding issue. o3 was a genuinely revolutionary model; if OpenAI had called it “GPT-5”, it would have met expectations. Instead, they called it “o3”, and called a minor incremental update a few months later “GPT-5”. Then people got mad that the exciting-sounding “GPT-5” was merely an incremental update. A secondary issue was that the router wasn’t very good, and so many queries got routed to a small version without thinking mode that was if anything a downgrade from o3. I think this tweet by Shakeel perfectly encapsulates the essence of GPT discourse in two sentences: …but maybe it’s worth asking why GPT-5 isn’t bigger than o3. Was 4.5 a failed attempt at scaling? Did it fail in a way that sort of back-handedly justifies the “lost steam” take? Does the answer depend on distinctions between pre-training scaling, post-training scaling, etc? How? 39: This month in etymology: did you know that “oy vey” is a “fully Germanic phrase” which is cognate with English “oh woe!” (h/t Wylfcen on X) 40: mRNA shows promise to be a game-changing treatment for cancer, but RFK is trying to halt research. But so far he can only starve it of money, not ban it, and the funding gap is only $500 million. Will there be enough philanthropic billionaires and private foundations to step up? Zvi points out that although there is usually a game of chicken where foundations are hesitant to touch something the government cancelled lest the government decide it can cancel everything and hope philanthropists pick up the bill, in this case there are no game theory considerations - RFK is halting it because he genuinely wants it halted, and they are thwarting him rather than playing into his hands. The only problem is that $500M is a lot of money for the private sector; a few foundations could technically afford it, but not many could afford it comfortably and still have money left over for the next few crises of this magnitude. I hope someone is trying to organize a coalition. 41: AI fantasy flash fiction Turing test. Eight stories about demons, four by famous fantasy authors, four by ChatGPT. After 3000 votes, AI wins: humans can't tell the difference and slightly prefer the AI stories. My own score was only 75%. But I will say that I thought Mark Lawrence's was obviously the best, I was ~100% sure it was human, and it convinced me that regardless of the official results it's still possible to write flash fiction that an AI obviously can't do. 42: “SignPro” offers customized “In This House We Believe” signs, try not to use this for evil. 43: China think tank assessment of how in control Xi is: still very in control, maybe not infinitely in control. 44: Related - did you know (h/t xlr8harder) that if you ask AI to write a science fiction story, it will very often name the protagonist “Elara Voss” (or some very close variant like Elena Voss), and this remains true across various models and versions? Related: Chelsea Voss of OpenAI is having a baby and has the opportunity to do the funniest thing. 45: “Hector (cloud) is a cumulonimbus thundercloud cluster that forms regularly nearly every afternoon on the Tiwi Islands in the Northern Territory of Australia…[he is sometimes called] Hector the Convector”. 46: British allergy sufferers who want to know the ingredients of things demand that British cosmetics stop listing their ingredients in Latin. “For example, sweet almond oil is Prunus Amygdalus Dulcis, peanut oil is Arachis Hypogaea, and wheat germ extract is Triticum Vulgare.” 47: Text-based RPG about being an NYT journalist at the Manifest prediction market conference. I make a brief appearance. 48: Study uses supposedly-random variation in doctor assignments to test whether the marginal mental health commitment is good or bad for patients, finds that it is quite bad. Freddie de Boer is violently skeptical (maybe literally so?) and makes some good points about how a single quasi-experimental study is never absolute proof. But I don’t think he quite justifies his opinion that the paper was irresponsible and should never have been published; it’s just a normal quasi-experimental study that we should nod and say “huh” at but not overweight as the culmination of all possible research that overcomes all possible priors. My prior is that the marginal commitment is pretty useless (many commitments are just “well, since this person arrived at our ED for some reason, it would look bad from a medico-legal perspective to just let them go, so let’s keep them a few days to evaluate” - and yeah, you should be upset about this) but I’m still surprised by how many outright negative (as opposed to zero) effects the researchers found. The strongest argument for negative effects is that it will make some people miss work and maybe lose their job. But this study found that commitment ~doubles the risk of near-term suicide (admittedly only from 1% to 2%), which would have been outside my confidence intervals for how bad it could be. I suspect confounding, but only on general principle, and I wouldn’t be too surprised either way. 49: This tweet is probably bait, but I found it a thought-provoking question: I think there’s a boring answer, where the law is more complex than just a single number and whatever kind of weird trafficking Epstein was doing is worse than whatever normal relationships these European laws are permitting. But assuming that there’s a substantive difference even after taking that into account, I think my answer is something like - we’ve got to divide kids from adults at some age, there’s a range of reasonable possible ages, we shouldn’t be too mad at other societies that choose different dividing lines within that range - but having decided upon the age, we’ve got to stick with it and take it seriously (in the sense of penalizing/shaming people who break it). This is more culturally relativist than I expected to find myself being, so good job to Richard for highlighting the apparent paradox. 50: Dilan Esper describes his experience as one of Hulk Hogan’s attorneys in the Gawker lawsuit (X). Parts I found interesting: none of the lawyers knew Thiel was funding the lawsuit; Gawker probably could have won if they had been slightly competent but kept "shooting themselves in the foot"; and Gawker probably could have won if they had just pixelated the private parts in the video. 51: Amazing concept and poems (link on X): I tried to see if AI could do this, and it did something that technically met the requirements but had zero artistic merit - using a lot of words like “nowhere” and “outside” in one, then separating them out to “no where” and “out side” in the other. I didn’t invest much energy in creating a clever prompt telling it not to do that, so feel free to report if you get better success. 52: New study claims consultants are actually good, at least for profits: "We find positive effects on labor productivity of 3.6% over five years, driven by modest employment reductions alongside stable or growing revenue" 53: A Polish team tries to test Peter Turchin’s equations for predicting political unrest on recent Polish history, has to make some changes but claims mostly positive results. 54: New big multi-author Substack, The Argument, trying to be a sort of center-left version of the model pioneered by The Free Press and other high-production-value ideological Substack properties. Excited to see Kelsey Piper is involved, and she starts off strong with a post on the latest round of First World basic income studies, which find few positive effects. This is surprising, because recipients didn’t waste the money on alcohol or gambling or anything - they paid down debt and got useful goods. Still, it didn’t even affect things that should have been obvious, like stress level. It’s not even clear that amounts of money large enough to help with rent made homeless people more likely to get houses! Matt Bruenig criticizes the article, accusing Kelsey’s studies of being downstream of Perry Preschool style dreams that exactly the right welfare program will have massively compounding effects that cut poverty out at the root and turn everyone into elite human capital; he thinks giving people money won’t do this, but it will increase equality and give the poor better lives. I assume he’s not a strong hereditarian, but his argument makes even more sense from that perspective, and I’ve certainly criticized dumb outcome measures like infant brain waves which we have only tenuous reasons to think are related to anything we care about. But Kelsey reasonably responds that the outcome measures she’s talking about include stress level and life satisfaction. To defuse this critique, Bruenig either has to argue that our construct “life satisfaction” doesn’t really measure whether someone’s life is satisfactory, or else claim that giving poor people satisfactory lives isn’t really what we’re going for - which I think would require more explanation on his part. There’s some further (impressively acrimonious) debate on X, but I don’t see anything that addresses my core concern. GiveDirectly, a charity involved in basic income experiments, has a presponse here; they say that some studies are positive, and that the ones that aren’t might have tried too little cash to matter, or been confounded by COVID making everything worse. They also point out that basic income is harder to study than traditional programs like giving people housing, because if you’re giving housing you can measure housing-related outcomes directly and have a pretty good chance of getting enough statistical power to find them, but since everyone spends cash on different things, the positive effects might be scattered across many different outcomes (and therefore too small to reach significance on each). Everyone involved in this debate wants to emphasize that the poor results are for First World studies only, and that studies continue to show large benefits to giving cash in the developing world. 55: Related: I was less impressed by The Argument’s first foray into housing policy, which follows an all-too-familiar pattern: Some people say they don’t like noise and disorder and try to make rules against it in their apartments.
Inline links: proposes a three-stage model of secularization, extraordinarily effective at teaching people golf, nxthompson on X, a huge survey, Steven Adler on AI psychosis, they cloned her, dozens of times, a lawsuit, Gwern, on X, got 40% of the e-commerce funding, What Happened To Pathology AI Companies?, Will Data Centers Crash The Economy?, Ruxandra Teslo provides the counterargument, and who is still fighting the good fight, countering Curtis Yarvin on the history of her native Romania, Ethan Mollick, Tyler Cowen, Nabeel Qureshi, Taelin, on formal benchmarks, speculated, a “reverse DeepSeek moment”, with Peter, this tweet by Shakeel, https://substackcdn.com/image/fetch/$s_!GJNZ!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ba0d8cf-fab8-4370-bcad-df789e157fdc_591x402.png, Wylfcen on X, Zvi points out that, AI fantasy flash fiction Turing test, customized “In This House We Believe” signs, China think tank assessment of how in control Xi is, xlr8harder, Chelsea Voss of OpenAI is having a baby, Hector (cloud), demand that British cosmetics stop listing their ingredients in Latin, Text-based RPG about being an NYT journalist at the Manifest prediction market conference, finds that it is quite bad, violently skeptical, literally so?, This tweet, https://substackcdn.com/image/fetch/$s_!S9fU!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa558c09b-7fb6-40a8-a8a0-27b658a2c876_576x687.png, describes his experience as one of Hulk Hogan’s attorneys in the Gawker lawsuit (X), link on X, https://substackcdn.com/image/fetch/$s_!zyh7!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F75e9f0f6-d794-4ea2-b24b-5d4803bf28dc_590x478.png, New study claims consultants are actually good, tries to test Peter Turchin’s equations for predicting political unrest on recent Polish history, The Argument, a post on the latest round of First World basic income studies, criticizes the article, infant brain waves, debate on X, has a presponse here, first foray into housing policy
…but maybe it’s worth asking why GPT-5 isn’t bigger than o3. Was 4.5 a failed attempt at scaling? Did it fail in a way that sort of back-handedly justifies the “lost steam” take? Does the answer depend on distinctions between pre-training scaling, post-training scaling, etc? How? 39: This month in etymology: did you know that “oy vey” is a “fully Germanic phrase” which is cognate with English “oh woe!” (h/t Wylfcen on X) 40: mRNA shows promise to be a game-changing treatment for cancer, but RFK is trying to halt research. But so far he can only starve it of money, not ban it, and the funding gap is only $500 million. Will there be enough philanthropic billionaires and private foundations to step up? Zvi points out that although there is usually a game of chicken where foundations are hesitant to touch something the government cancelled lest the government decide it can cancel everything and hope philanthropists pick up the bill, in this case there are no game theory considerations - RFK is halting it because he genuinely wants it halted, and they are thwarting him rather than playing into his hands. The only problem is that $500M is a lot of money for the private sector; a few foundations could technically afford it, but not many could afford it comfortably and still have money left over for the next few crises of this magnitude. I hope someone is trying to organize a coalition. 41: AI fantasy flash fiction Turing test. Eight stories about demons, four by famous fantasy authors, four by ChatGPT. After 3000 votes, AI wins: humans can't tell the difference and slightly prefer the AI stories. My own score was only 75%. But I will say that I thought Mark Lawrence's was obviously the best, I was ~100% sure it was human, and it convinced me that regardless of the official results it's still possible to write flash fiction that an AI obviously can't do. 42: “SignPro” offers customized “In This House We Believe” signs, try not to use this for evil. 43: China think tank assessment of how in control Xi is: still very in control, maybe not infinitely in control. 44: Related - did you know (h/t xlr8harder) that if you ask AI to write a science fiction story, it will very often name the protagonist “Elara Voss” (or some very close variant like Elena Voss), and this remains true across various models and versions? Related: Chelsea Voss of OpenAI is having a baby and has the opportunity to do the funniest thing. 45: “Hector (cloud) is a cumulonimbus thundercloud cluster that forms regularly nearly every afternoon on the Tiwi Islands in the Northern Territory of Australia…[he is sometimes called] Hector the Convector”. 46: British allergy sufferers who want to know the ingredients of things demand that British cosmetics stop listing their ingredients in Latin. “For example, sweet almond oil is Prunus Amygdalus Dulcis, peanut oil is Arachis Hypogaea, and wheat germ extract is Triticum Vulgare.” 47: Text-based RPG about being an NYT journalist at the Manifest prediction market conference. I make a brief appearance. 48: Study uses supposedly-random variation in doctor assignments to test whether the marginal mental health commitment is good or bad for patients, finds that it is quite bad. Freddie de Boer is violently skeptical (maybe literally so?) and makes some good points about how a single quasi-experimental study is never absolute proof. But I don’t think he quite justifies his opinion that the paper was irresponsible and should never have been published; it’s just a normal quasi-experimental study that we should nod and say “huh” at but not overweight as the culmination of all possible research that overcomes all possible priors. My prior is that the marginal commitment is pretty useless (many commitments are just “well, since this person arrived at our ED for some reason, it would look bad from a medico-legal perspective to just let them go, so let’s keep them a few days to evaluate” - and yeah, you should be upset about this) but I’m still surprised by how many outright negative (as opposed to zero) effects the researchers found. The strongest argument for negative effects is that it will make some people miss work and maybe lose their job. But this study found that commitment ~doubles the risk of near-term suicide (admittedly only from 1% to 2%), which would have been outside my confidence intervals for how bad it could be. I suspect confounding, but only on general principle, and I wouldn’t be too surprised either way. 49: This tweet is probably bait, but I found it a thought-provoking question: I think there’s a boring answer, where the law is more complex than just a single number and whatever kind of weird trafficking Epstein was doing is worse than whatever normal relationships these European laws are permitting. But assuming that there’s a substantive difference even after taking that into account, I think my answer is something like - we’ve got to divide kids from adults at some age, there’s a range of reasonable possible ages, we shouldn’t be too mad at other societies that choose different dividing lines within that range - but having decided upon the age, we’ve got to stick with it and take it seriously (in the sense of penalizing/shaming people who break it). This is more culturally relativist than I expected to find myself being, so good job to Richard for highlighting the apparent paradox. 50: Dilan Esper describes his experience as one of Hulk Hogan’s attorneys in the Gawker lawsuit (X). Parts I found interesting: none of the lawyers knew Thiel was funding the lawsuit; Gawker probably could have won if they had been slightly competent but kept "shooting themselves in the foot"; and Gawker probably could have won if they had just pixelated the private parts in the video. 51: Amazing concept and poems (link on X): I tried to see if AI could do this, and it did something that technically met the requirements but had zero artistic merit - using a lot of words like “nowhere” and “outside” in one, then separating them out to “no where” and “out side” in the other. I didn’t invest much energy in creating a clever prompt telling it not to do that, so feel free to report if you get better success. 52: New study claims consultants are actually good, at least for profits: "We find positive effects on labor productivity of 3.6% over five years, driven by modest employment reductions alongside stable or growing revenue" 53: A Polish team tries to test Peter Turchin’s equations for predicting political unrest on recent Polish history, has to make some changes but claims mostly positive results. 54: New big multi-author Substack, The Argument, trying to be a sort of center-left version of the model pioneered by The Free Press and other high-production-value ideological Substack properties. Excited to see Kelsey Piper is involved, and she starts off strong with a post on the latest round of First World basic income studies, which find few positive effects. This is surprising, because recipients didn’t waste the money on alcohol or gambling or anything - they paid down debt and got useful goods. Still, it didn’t even affect things that should have been obvious, like stress level. It’s not even clear that amounts of money large enough to help with rent made homeless people more likely to get houses! Matt Bruenig criticizes the article, accusing Kelsey’s studies of being downstream of Perry Preschool style dreams that exactly the right welfare program will have massively compounding effects that cut poverty out at the root and turn everyone into elite human capital; he thinks giving people money won’t do this, but it will increase equality and give the poor better lives. I assume he’s not a strong hereditarian, but his argument makes even more sense from that perspective, and I’ve certainly criticized dumb outcome measures like infant brain waves which we have only tenuous reasons to think are related to anything we care about. But Kelsey reasonably responds that the outcome measures she’s talking about include stress level and life satisfaction. To defuse this critique, Bruenig either has to argue that our construct “life satisfaction” doesn’t really measure whether someone’s life is satisfactory, or else claim that giving poor people satisfactory lives isn’t really what we’re going for - which I think would require more explanation on his part. There’s some further (impressively acrimonious) debate on X, but I don’t see anything that addresses my core concern. GiveDirectly, a charity involved in basic income experiments, has a presponse here; they say that some studies are positive, and that the ones that aren’t might have tried too little cash to matter, or been confounded by COVID making everything worse. They also point out that basic income is harder to study than traditional programs like giving people housing, because if you’re giving housing you can measure housing-related outcomes directly and have a pretty good chance of getting enough statistical power to find them, but since everyone spends cash on different things, the positive effects might be scattered across many different outcomes (and therefore too small to reach significance on each). Everyone involved in this debate wants to emphasize that the poor results are for First World studies only, and that studies continue to show large benefits to giving cash in the developing world. 55: Related: I was less impressed by The Argument’s first foray into housing policy, which follows an all-too-familiar pattern: Some people say they don’t like noise and disorder and try to make rules against it in their apartments.
Inline links: Wylfcen on X, Zvi points out that, AI fantasy flash fiction Turing test, customized “In This House We Believe” signs, China think tank assessment of how in control Xi is, xlr8harder, Chelsea Voss of OpenAI is having a baby, Hector (cloud), demand that British cosmetics stop listing their ingredients in Latin, Text-based RPG about being an NYT journalist at the Manifest prediction market conference, finds that it is quite bad, violently skeptical, literally so?, This tweet, https://substackcdn.com/image/fetch/$s_!S9fU!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa558c09b-7fb6-40a8-a8a0-27b658a2c876_576x687.png, describes his experience as one of Hulk Hogan’s attorneys in the Gawker lawsuit (X), link on X, https://substackcdn.com/image/fetch/$s_!zyh7!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F75e9f0f6-d794-4ea2-b24b-5d4803bf28dc_590x478.png, New study claims consultants are actually good, tries to test Peter Turchin’s equations for predicting political unrest on recent Polish history, The Argument, a post on the latest round of First World basic income studies, criticizes the article, infant brain waves, debate on X, has a presponse here, first foray into housing policy