Herasight

Article

Herasight is a recurring organization in the Astral Codex Ten archive, appearing 3 times across 3 issues between July 31, 2025 and December 10, 2025. The archive places it in contexts such as “Herasight’s numbers on how breast cancer risk goes down”; “Genomic Prediction and Herasight try to help by providing semi-objective measures”; “Six extra IQ points (Herasight’s estimate with five embryos)“. It most often appears alongside 23andme, ACX, Anthropic.

Metadata

Category: Organizations
Mention count: 3
Issue count: 3
First seen: July 31, 2025
Last seen: December 10, 2025

Appears In

- 23andme (2 shared issues)
- ACX (2 shared issues)
- Anthropic (2 shared issues)
- Genomic Prediction (2 shared issues)
- Nucleus (2 shared issues)
- Orchid (2 shared issues)
- Substack (2 shared issues)
- UK (2 shared issues)
- US (2 shared issues)
- 100 Above The Park (1 shared issues)
- 2024 Post-Mortem Of Neoreaction (1 shared issues)
- 23andMe (1 shared issues)

External Links

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.

Suddenly, Trait-Based Embryo Selection

July 31, 2025 · Original source

Sample Nucleus results. And this week, Herasight4 entered the space with the most impressive disease risk scores yet, an IQ predictor worth 6-95 extra points, and a series of challenges to competitors, whom they call out for insufficient scientific rigor. Their most scathing attack is on Nucleus itself, accusing its predictions of being misleading and unreliable. Let’s start with the science, then move on to the companies and see if we can litigate their dispute. In Theory, All Of This Should Work Polygenic embryo screening is a natural extension of two well-validated technologies: genetic testing of embryos, and polygenic prediction of traits in adults. Genetic testing of embryos has been done for decades, usually to detect chromosomal abnormalities like Down Syndrome or simple single-gene disorders like cystic fibrosis. It’s challenging - you need to take a very small number of cells (often only 5-10) from a tiny proto-placenta that may not have many cells to spare, and extract a readable amount of genetic material from this limited sample - but there are known solutions that mostly work. But most traits are polygenic, requiring information about thousands or tens of thousands of genes to predict. These are too complicated to understand fully at current levels of technology, but some studies have chipped away at the problem and gotten a partial understanding. Often this looks like being able to predict a few percent of the variance in a trait, and determine whether someone’s genetic risk is slightly higher or lower than average. Polygenic prediction of traits in adults is still young and full of hidden pitfalls. Last month, we discussed how some early studies unknowingly conflated direct genetic effects and various confounders6 - for example, they tended to pick up on genes associated with well-off ethnic groups or families who had good health outcomes for social reasons. Pinpointing the direct component requires an additional step where researchers validate their algorithms within families (for example, on pairs of siblings where one has a higher polygenic score than the other) to see how much predictive power remains. This is especially important for embryo selection companies, whose entire value proposition depends on comparing two genomes from the same family. How have they done? It depends on the number of embryos they have to work with; the more embryos, the better you can do by selecting the best. Herasight’s numbers on how breast cancer risk goes down with number of embryos used in selection. A typical round of IVF produces 1-10 embryos (younger women usually = more). Women with polycystic ovarian syndrome (prevalence: 10%) may get as many as 20. For more, you will probably need to do multiple IVF rounds. Here is a table of different companies’ reported risk reductions, slightly adjusted7 for different reporting conventions but otherwise taking all claims at face value (we’ll talk about how wise that is later). Relative risk reduction for five conditions (gray = no data / disputed data). Here baseline is for embryos neither of whose parents have the condition. GP and Orchid both say their technology has improved since reporting these numbers and they will report better numbers soon. GP numbers are not within-family validated and might be lower if they were. Absolute risk after selection for five conditions (gray = no data / disputed data), ibid. Some people might genuinely want to select on a single condition. For example, people with a strong family history of schizophrenia might want to minimize the chance of their children getting the disease; for these people, reducing schizophrenia risk by 58% (while keeping everything else constant) sounds pretty good. Everyone else probably wants a generically healthy embryo with low risk of all conditions. Exactly how this works depends on the customer’s own values - would they prefer an embryo with lower cancer risk to one who will have fewer heart attacks? - and the exact benefits will depend on how parents make that decision. Genomic Prediction and Herasight try to help by providing semi-objective measures of which embryo is overall healthiest according to different conditions’ effects on longevity and patient-rated quality of life. For Genomic Prediction, that’s the “embryo health score” If you selected the single highest-health-score embryo from a set of five, here’s how they’d do: For Herasight, it’s a “polygenic longevity index”. They don’t give exact risk reduction numbers for each disease, saying that it depends too much on a couple’s specific family history, but say that most people gain 1-4 years of healthy life (when I test it on a set of twenty embryos, the the healthiest gets an extra 1.66 years). How much would you pay to give your children an extra 1-4 years of healthy life? This is no longer a hypothetical question. Here are the costs of the companies in this space: Is it worth it? If: You’re already doing IVF

Inline links: Herasight, 4, 5, 6, https://substackcdn.com/image/fetch/$s_!VOdq!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb1ba32f3-72fa-4be1-846c-6b0b04a5a213_774x279.png, 7, https://substackcdn.com/image/fetch/$s_!0oUh!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe8419603-9239-43bb-8c79-77b078ff0789_548x136.png, https://substackcdn.com/image/fetch/$s_!rpEJ!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc539a717-a130-460d-90c9-4ab64619f26d_548x133.png, https://substackcdn.com/image/fetch/$s_!3Kc6!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbda325bb-13fb-4c27-b8c3-24facce5c71a_676x153.png, https://substackcdn.com/image/fetch/$s_!t1Am!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F831ab3c6-4053-4ff9-bc2f-879aee4349cf_673x740.png, https://substackcdn.com/image/fetch/$s_!Q2vE!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fadcf2bb4-8dd1-4a88-9728-6953a820971b_422x575.png

Herasight’s numbers on how breast cancer risk goes down with number of embryos used in selection. A typical round of IVF produces 1-10 embryos (younger women usually = more). Women with polycystic ovarian syndrome (prevalence: 10%) may get as many as 20. For more, you will probably need to do multiple IVF rounds. Here is a table of different companies’ reported risk reductions, slightly adjusted7 for different reporting conventions but otherwise taking all claims at face value (we’ll talk about how wise that is later). Relative risk reduction for five conditions (gray = no data / disputed data). Here baseline is for embryos neither of whose parents have the condition. GP and Orchid both say their technology has improved since reporting these numbers and they will report better numbers soon. GP numbers are not within-family validated and might be lower if they were. Absolute risk after selection for five conditions (gray = no data / disputed data), ibid. Some people might genuinely want to select on a single condition. For example, people with a strong family history of schizophrenia might want to minimize the chance of their children getting the disease; for these people, reducing schizophrenia risk by 58% (while keeping everything else constant) sounds pretty good. Everyone else probably wants a generically healthy embryo with low risk of all conditions. Exactly how this works depends on the customer’s own values - would they prefer an embryo with lower cancer risk to one who will have fewer heart attacks? - and the exact benefits will depend on how parents make that decision. Genomic Prediction and Herasight try to help by providing semi-objective measures of which embryo is overall healthiest according to different conditions’ effects on longevity and patient-rated quality of life. For Genomic Prediction, that’s the “embryo health score” If you selected the single highest-health-score embryo from a set of five, here’s how they’d do: For Herasight, it’s a “polygenic longevity index”. They don’t give exact risk reduction numbers for each disease, saying that it depends too much on a couple’s specific family history, but say that most people gain 1-4 years of healthy life (when I test it on a set of twenty embryos, the the healthiest gets an extra 1.66 years). How much would you pay to give your children an extra 1-4 years of healthy life? This is no longer a hypothetical question. Here are the costs of the companies in this space: Is it worth it? If: You’re already doing IVF

Inline links: 7, https://substackcdn.com/image/fetch/$s_!0oUh!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe8419603-9239-43bb-8c79-77b078ff0789_548x136.png, https://substackcdn.com/image/fetch/$s_!rpEJ!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc539a717-a130-460d-90c9-4ab64619f26d_548x133.png, https://substackcdn.com/image/fetch/$s_!3Kc6!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbda325bb-13fb-4c27-b8c3-24facce5c71a_676x153.png, https://substackcdn.com/image/fetch/$s_!t1Am!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F831ab3c6-4053-4ff9-bc2f-879aee4349cf_673x740.png, https://substackcdn.com/image/fetch/$s_!Q2vE!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fadcf2bb4-8dd1-4a88-9728-6953a820971b_422x575.png

You’re okay using expected utility calculations where a 50% chance of preventing X is half as good as fully preventing X. …then I’ll go out on a limb and say yeah, obviously it’s worth it. Consider e.g. Genomic Prediction, which costs $3,250 for five embryos and claims to lower absolute risk of Type 2 diabetes by 12%. That implies that not getting Type 2 diabetes is worth $27,000. Ask anybody dealing with regular insulin injections (let alone limb amputations) whether it would be worth $27,000 to wave a magic wand and not have Type 2 diabetes! It’s not a hard question! And that’s just one of a dozen conditions you can lower the risk for! Other ones, like not getting breast cancer, might be so valuable that it’s hard to even attach numbers! (but maybe the low time discount rate is a mistake? Suppose you invest the $3,250 in an index fund that makes 7% over inflation, then give it to your future child when they turn 45 (average age of type 2 diabetes diagnoses). Now it’s worth $75,000. Is this the “true” cost of the intervention? Does it matter that this counterfactual is fake and most people don’t do this?) What about IQ? Six extra IQ points (Herasight’s estimate with five embryos) is about a quarter of the gap between the average person and the average Ivy League student. The benefits of intelligence are hard to quantify, but it’s been shown to have probably-causal positive effects on income, mortality, and achievement. Probably the income effects alone make up for the cost of intervention - again assuming total parent-child altruism and low discount rate8. So if we accept all of these claims and assumptions, the choice seems obvious. It’s probably even obvious for governments to pay for all citizens to get these, given how much they’d save on health care costs. In Practice, It’s Complicated Critics have raised both scientific and ethical objections to polygenic embryo screening. Most significantly, it’s been condemned by various bodies including the Society For Psychiatric Genetics, the European Society of Human Genetics, and the Behavioral Genetics Society. Their statements are . . . not good. They tend towards vague language about how people are more than just their genes, or how no genetic test can be perfect, or how embryo screening is not exactly the same thing as some other form of screening which has a longer history and more proponents. “Although in general higher scores mean you are more likely to have a condition, many healthy people will have high scores; others might develop the condition even with a low score”, says the Society for Psychiatric Genetics, as if they have just blown the lid off some dastardly conspiracy. “Screening embryos for psychiatric conditions may increase stigma surrounding these diagnoses”, they continue - an objection which, taken seriously, could be used to ban every form of medical treatment. We will mostly ignore these people and try to imagine the objections that mildly competent critics might raise, some of which will coincidentally overlap with the content of the non-hypothetical statements. Scientific Objection: Efficacy Are we sure this works at all? A typical polygenic score is created by collecting thousands or millions of adult genomes, then matching genetic information with surveys about who has the trait/condition of interest. Reputable studies then test these scores on holdout samples - adults who were not used to make the score, to see if they still accurately predict who has the trait/condition. Polygenic embryo selection depends on an assumption that the scores which work in these kinds of retrospective tests will also work prospectively on embryos. This assumption hasn’t been formally proven in studies (which would require years to decades to conduct), but seems common-sensical. The strongest challenge to the application of polygenic scores for embryo selection comes from a recent body of research showing that most scores combine causal genetic effects with population stratification, and therefore can be expected to lose much of their predictive power when comparing two members of the same family (e.g. two embryos from the same couple). There is increasing agreement in the field that unless scores are validated within families, headline results like “decreases risk of X by Y%” will be large overestimates. When I talked to company representatives, they all said that they took accuracy extremely seriously and had various white papers and journal articles where anyone could double-check their methodology. But I attended an industry conference a few months ago, and the gossip level was comparable to a high school cafeteria (minus the sex rumors - most of the attendees were having their own kids via IVF). Everyone had some story about someone being careless or fudging their numbers. Some of the conflicts broke out into the open on Wednesday, when Herasight left stealth and published a white paper and associated blog post. They criticize Genomic Prediction for reporting between-family rather than within family results9, and Orchid for smuggling a term for age into their Alzheimer’s predictor (unsurprisingly, this makes it work better). We’ll get to their accusations against Nucleus below. Note that this was recent enough that competitors haven’t had time to respond or to air their own criticisms of Herasight; if this happens, I’ll try to keep you updated. Maybe this is cope, but my optimistic perspective is that this bounds the damage. This obviously isn’t a field capable of maintaining a conspiracy of silence. But aside from the Nucleus allegations, the complaints aren’t existential. Maybe some numbers are too high, maybe some predictors are slightly rigged. But the more we learn about these admittedly concerning problems, the more we can hope that we’d have heard about it if nothing worked at all. Overall my strongest opinion on the scientific criticisms is: Authorities on all sides have cited Alex Young10 as an authority on how polygenic scores can be confounded or misleading.

Inline links: income, mortality, achievement, 8, Society For Psychiatric Genetics, European Society of Human Genetics, Behavioral Genetics Society, white paper, blog post, 9, 10

Open Thread 394

August 11, 2025 · Original source

4: In the post on embryo selection, I mentioned that Herasight criticized Orchid's Alzheimer's predictor. A representative of Orchid reached out to say they stood by their methodology:

Inline links: the post on embryo selection, criticized

Herasight seems to be misreading our whitepaper. The “Performant Alzheimer’s disease risk stratification” section is meant to show the kind of performance patients can expect—people in the top 5% have an OR of 5.80, top 3% is 7.35, and top 1% is 11.69. This odds ratio is what is used to present embryo disease risk to patients and does not include covariates. The “Comparison to Published Benchmarks” section is just about comparing our models to others in the literature. To allow a head to head comparison, we used the same metric (AUC) and covariates as the paper we’re comparing against. However, to avoid future confusion, we’ve just added a sentence clarifying the AUC without covariates (0.724).

Inline links: our whitepaper

They also state that Herasight, like themselves, has only validated the predictors where there’s enough data (e.g. not schizophrenia), and they object to Herasight claiming superiority in this area.

Links For December 2025

December 10, 2025 · Original source

Yes, voters react positively both to candidates “over the age of 50” and candidates “under the age of 50”. Just don’t run 50 year olds! 16: I previously blogged about how embryo-selection company Nucleus appeared scammy. Sichuan_Mala looks deeper and agrees they seems scammy. Besides what I found, she finds several errors in the white paper, apparently fake customer reviews, and an accusation of IP theft from competitor Genomic Prediction. She also accuses them of plagiarizing competitor Herasight’s work, although it’s a bit subtle and I don’t know enough about field norms to know whether this is a case of flattery-by-imitation or totally out of bounds. A Nucleus researcher responds to the scientific allegations here, saying that the “plagiarism” was just convergent methodologies. And Nucleus CEO Kian Sadeghi goes on the TBPN podcast here to rebut the business allegations, saying that the customer reviews are real although some photos were changed for privacy reasons. There’s an appearance/facedox by fellow Nucleus skeptic Cremieux Recueil, although Kian declines to debate him directly; you can see Cremieux’s postmortem of the episode here. My opinion is that as potential customers, you are under no obligation to care whether the company plagiarizes papers or fakes reviews, but you should care about whether their genetic tests are good, and I continue to think they’re not. Their old competitor Genomic Prediction is cheaper, and their new competitor Herasight has more powerful predictors, so you’re excused from having to have an opinion on this, and should just use someone else’s product. Related: Gene Smith’s rundown of the pros and cons of every company in the embryo selection space (X). 17: And related: a Herasight client describes her experience with embryo selection, and her feelings upon the birth of her selected child. 18: Lars Doucet, guest author of several ACX posts on Georgism, reviews The Land Trap by Mike Bird. “Land is a big deal, and always has been. [But] land has only recently been financialized. Financializing land causes ‘the land trap’ . . . [where] land slowly sucks up all your economy’s productivity, inflating a dangerous real estate bubble that eventually pops, leaving disaster in its wake”. Also, “Fiat currency isn’t backed by nothing, as commonly supposed, but by land.” 19: New research analyzes Hitler’s DNA. Findings: he had Kallman Syndrome, a rare disorder of sexual development associated with low testosterone, micropenis, and small testicles (ironically, the WWII song about Nazi sexual inadequacies only accuses Goering and Himmler of this, but lets Hitler off). Contra galaxy-brained rumors, he did not have any Jewish ancestry. And he had “very high scores - in the top one percent - for a predisposition to autism, schizophrenia and bipolar disorder”. When I wrote this post, a reader asked me what it would look like for someone to have high propensity for both autism and schizophrenia at the same time. Well . . . 20: The wealth of cities (h/t @StatisticUrban): 21: Update on Tech PACs Are Closing In On The Almonds: pro-AI safety politician Alex Bores announced his candidacy for Congress in New York. As expected, the A16Z pro-AI PAC announced a “multibillion dollar effort to sink [his] campaign” (wait, multi-billion on one candidate? is that a typo?) This doesn’t seem to be going very well for them so far. Bores has masterfully leveraged (X) the unprecedented opposition from Big Tech into a selling point. …and raised $1.2 million on his first day, breaking fundraising records (I was told this was because of pro-AI-safety EAs, but others credit AIPAC and the Israel lobby). And most recently, Jami Floyd, one of Bores’ opponents and a possible beneficiary of anti-Bores spending, has condemned it (X) and demanded that the AI industry stop trying to help her. Impressive work from everybody. Related: New $50 million pro-AI-regulation SuperPAC, I assume EA-linked but have no special knowledge. 22: Related: Pre-emption is when Congress blocks states from making legislation on a topic, saying it will decide all the laws itself. The states have signaled willingness to regulate AI pretty hard, so Big Tech has been pushing for AI pre-emption to (in their opinion) prevent an overly complicated patchwork of regulations, or (in their opponents’ opinion) shift everything to a Republican Congress that will drop the ball on regulation entirely. After their first attempt in June was defeated by a coalition of anti-tech liberals and anti-tech conservatives, we discussed (1, 2) the effort by moderates on both sides to create a compromise proposal which pre-empted state laws but guaranteed good federal regulation on important topics. The most recent news is that extremists sidelined the moderates and tried to slip a hardline preemption deal with no compromises into the National Defense Authorization Act, a defense budget bill which is notoriously secretive and hard for the public to learn about. This didn’t work; some of the same coalition, plus a group of Republican state legislators including Ron DeSantis, pressured the GOP to drop it. The next battleground is a potential Trump executive order; although Trump cannot constitutionally ban states from regulating AI, he will threaten them with various consequences like lawsuits or withdrawal of federal funding. The buzz in the policy circles I’m in is that this might backfire; blue state politicians love starting fights with Trump in order to look tough to their blue state electorates. No, no, please don’t give me headlines like “TRUMP CONDEMNS GAVIN NEWSOM FOR TRYING TO PROTECT CALIFORNIA’S CHILDREN FROM AI SLOP”! Anything but that! 23: Related: Trump has decided to sell some of America’s best AI chips to China, supercharging their AI development and crippling ours. The most charitable read is that his administration doesn’t really believe AI matters so they think it’s fine to forfeit it for short-term gain; the least charitable that it’s downstream of the companies involved paying Trump enormous bribes in hopes of exactly this outcome . We’re headed for the dumbest possible world, where we sacrifice our chance to thoughtfully address AI’s social impacts because “tHaT wOuLd mAkE uS lOsE tHe rAcE wItH ChInA”, then throw away the race with China in one fell swoop by handing them our technology for no reason. Shame on everyone involved, especially the people who shout over any discussion of safety with “bUt ChInA” yet have stayed totally silent about this. Our best hope now is that China refuses the chips, either because they want to privilege their own tech companies, or because they think we can’t possibly be this stupid and it must be some kind of spy plot. 24: Related: how the American public’s opinions on AI are changing (from David Shor, h/t Daniel Eth on X): If this is to be taken seriously, AI is already a bigger political issue than abortion, climate change, or the environment. I fail my 2023 prediction that there was only a 20% chance this would happen by 2028. 25: Related: Bernie Sanders in The Guardian: “There is a very real fear that, in the not-so-distant future, a super-intelligent AI could replace humans in controlling the planet.” The Left has a complicated relationship with existential risk from AI: they really hate AI, which in theory should push them towards yet another reason to be against it. But they hate AI so much that they need to believe every negative thing about it at the same time, and one of those negative things is that it’s just a scam and will never work, and this naturally pushes against being concerned about x-risk. But as AI improves, will the “just a scam” position become less tenable, shunting the associated psychic energy into other reasons to hate AI (including x-risk concerns)? 26: Qualia Research Institute has released a video describing some of the work they’ve been doing the past year - The Oscilleditor: An Algorithmic Breakthrough for Psychedelic Visual Replication (1080p•⚠️SEIZURE): 27: Jesse Arm (X): “A majority of American rabbinical students are now women. Most are also LGBTQ. That includes Modern Orthodoxy. Remove Modern Orthodoxy and the numbers climb even higher.” Clergy have always served as spiritual counselors; as religions liberalize and other roles become less important, the therapist role starts to predominate. But 75% of therapists in the US are female; at the limit of liberalization where clergyman = therapist, we should expect the same gender ratio. 28: The latest news on the COVID origins debate: scientists find a naturally-occuring bat coronavirus with a COVID-like furin cleavage site. This is a point in favor of the natural origins hypothesis, since the second-best argument for lab leak was that COVID’s furin cleavage site was too strange to evolve naturally. But I think arguments that lab leak has “fallen apart” are premature: the best argument (COVID emerged only a few miles from the biggest coronavirus gain-of-function lab in the Eastern Hemisphere) remains strong. I update from something like 95% chance it’s natural to something like 96%, but not 99.99% or anything. And here’s a lab leaker arguing that COVID’s furin cleavage site is out-of-frame and so still more unnatural-looking than the one on the recently-discovered bat virus. 29: Nicholas Decker (econ blogger, famous for his controversial autistic takes and Secret Service visit) has a dating doc. Most interesting section is the one about children: he wants to have them, but doesn’t think they should be genetically related to him. From here: If this appeals to you, you can find his contact info on the document. Related: Governor Jared Polis of Colorado is a fan of Nicholas Decker and Richard Hanania. 30: Matt Yglesias comes out as aphantasic (unable to see images in his “mind’s eye”). He says that contra the usual perspective that frames this as a deficit, he finds it helpful. For example, once he got assaulted, and he remembers on an intellectual level that it happened, but since “I wasn’t taking pictures of myself getting kicked in the head so, as far as I’m concerned, it’s like it happened to someone else” (Matt usually has good instincts, so I’m surprised he uses an example which will be such catnip to his conservative critics). He thinks it makes him a better reasoner / statistics blogger / effective altruist to be able to “get a statistically valid view of the situation, not overindex on the happenstance of your life.” For what it’s worth, I’ll give my contrary data point - I think of myself as a reasoner / statistics blogger / effective altruist in a pretty similar vein as Matt, but AFAICT my visual imagination is totally normal; if other people are having their emotions yanked around by vivid images, that’s a skill issue. 31: Lakshya Jain in The Argument: The COVID political backlash [to the Democratic Party] has disappeared. Despite the narrative, polls show that voters don’t favor or disfavor either party over COVID, mostly still think school closures were necessary, and are about evenly split on vaccine mandates. I guess I can’t disagree with this poll - it seems well-done - but I still wonder whether something is being missed. Maybe it didn’t make the ~50% of voters who are naturally liberal desert the cause, but it energized conservatives in a way that might otherwise not have happened? Related, from Rob Wiblin on X, on balance Britons think the government response to COVID was not strict enough. 32: Related: Back when neoreaction was a big deal, I occasionally discussed posts by neoreactionary blogger Spandrell of Bloody Shovel. If you’re wondering what happened to him, you can read his 2024 Post-Mortem Of Neoreaction here, where he discusses how he fell out of love with the movement (warning: he has not fallen out of love with racial slurs). As a former fascist sympathizer, I can see why [fascism is on the downswing]. The allure of fascism in 2024 is much, much diminished. For a few reasons. A big one was COVID. See, the point of fascism is that Collective Action is necessary to have nice things. We need a strong government committed to the good of the people. Yarvin showed his preference early when he started his new Substack by quoting Cicero’s phrase “Salus populi suprema lex”. The health of the people is the most important law. Cicero wasn’t a fascist of course, nor is Yarvin really; a big point of fascism is to narrowly define the populus as an ethnic group with demonstrable ties to blood. That makes the government’s ties to the people stronger, increasing their commitment to do Good Collective Action. Which is important. Very important. A lot of good things can come of intelligently done Collective Action. Fascist Italy made the trains run on time. Nazi Germany fixed the terrible Weimar economy. East Asian countries are all effectively fascist states, if with less ideological baggage (yellows just aren’t like that), and they are all nice, clean, safe places with healthy economies. Fascism is not a panacea but it works, when you let it. Strong government can be pretty neat. So why is strong government less appealing these days? Well, COVID happened. And our governments were pretty damn strong in dealing with it. They made strong laws and enforced them. And what did they do with their power? Absolutely retarded shit. They destroyed the world economy and made 95% of people completely miserable for 18 months. Up to 3 long years in some places. Again, as an Orient enjoyer I was very sympathetic of strong effective government. My life has been pretty cozy thanks to it for the past decades. But after seeing boomers, hypochondriacs, and menopausal women take the reins and use it against healthy people, I’m fucking done with strong effective government. Fuck that shit, I’m out. I don’t want to see strong effective government ever again. I was very lucky that I was out of China in November 2019. It was a fluke really. I moved to the Golden Triangle after that and the law of the jungle was much, much nicer during the Doctors Plague of 2020-2022. But I spent a few months in Europe during the time and man, that was brutal. Not just seeing how retarded governments were; the level of compliance by the people was so disheartening. Imagine being a sincere fascist and seeing your people behave like that. These are my people? My Volk? Am I supposed to sacrifice life and limb for the salus of this populus? Fuck that. Let them cook, they deserve everything that’s coming to them [...] Is there a way to make the body healthy again? I do think so. I think there’s still place for a successor right wing ideology which is neither Christian fundamentalism or robot worship. And it will happen; but it won’t happen on Twitter. Maybe it can happen on Urbit, or right here in this site. I have some ideas myself, and I invite you to join me and build this together. It would be funny if the solution to the paradox Jain highlights was that for every time a COVID lockdown turned a liberal into a conservative, it turned one fascist into a moderate, for a net rightward shift of zero. 33: Also from an Argument poll: In a hypothetical Presidential matchup, Gavin Newsom beats JD Vance 54-46. I’m split between the usual heuristic of ignoring any polling more than a year before an election, and the fact that this is a remarkably big lead for polarized 21st century America. 34: Jerl wades into the David Hume on miracles debate. 35: AI Teddy Bears: A Brief Investigation. The good news is that your child’s AI teddy bear is hard to jailbreak and probably will not tell them where to find guns: The other good news is that somehow they don’t charge a subscription, which makes them a way to get usually-subscription-only AI models for free. How is this possible? “[The most likely hypothesis is that] Witpaw is an adorable piece of spyware and he’s selling my data to the CCP”. 36: This month’s anti-people-named-Sacks content: NYT on Trump AI czar David Sacks’ conflicts of interest; New Yorker on whether neurologist Oliver Sacks used his case studies to work through his own issues rather than presenting them accurately. [EDITED TO ADD: I originally framed it this way as a joke, but on further research I think David and Oliver are related. Wikipedia says that Oliver was first cousins with Israel statesman Abba Eban, and that Abba Eban was born to Lithuanian Jewish parents in Cape Town. David Sacks’ bio says he was born to Jewish parents in Cape Town, and this article specifies that they were Lithuanian. I doubt there were too many Lithuanian Jewish families named Sacks in mid-1900s Cape Town, so sure, related!) 37: Orca Sciences: There Has To Be A Better Way To Make Titanium. Titanium is a great metal - strong, light, and tough. If we had cheap titanium, it could revolutionize manufacturing the way cheap steel and aluminum did in previous eras. So why don’t we? Not because titanium is rare: it’s “the 9th most common element in the earth’s crust”. Rather, it’s very complicated and expensive to extract from its ore. Some kind of breakthrough in titanium extraction processes always seems tantalizingly close, but has never quite materialized. Is there any hope? 38: If Asians Are Lactose Intolerant, Why All The Milk Tea? Lactose intolerance has confused me for a long time - 23andMe tells me that I’m lactose intolerant, but I drink milk regularly without problems, so what’s up? This post’s answer: lactose-intolerant people who don’t usually drink milk will get sick if they start suddenly. Lactose-intolerant people who drink milk regularly since childhood develop gut microbiota that can digest milk, but which demand an expensive “tax” in calories. Lactose-tolerant people will always be able to digest milk and absorb all the calories themselves. 39: How do different majors change college students’ political beliefs? No surprise that the humanities and social sciences shift people left; no surprise that business and economics shift them right. I was a little surprised that engineering shifts people right a little, and that Education of all things shifts people right (albeit only slightly). How is that even possible? Are these people coming in as Mao Zedong and leaving as “only” Leon Trotsky? Also, Political Science is exactly neutral, lol. [EDIT: I misunderstood, they’re using natural sciences as a zero point, this is a reasonable choice but slightly changes the interpretation] 40: Kindkristin: Language models improved my mental health. 41: More floor employment, from the WSJ (h/t @LaocoonofTroy): Big Paychecks Can’t Woo Enough Sailors For America’s Commercial Fleet: “Straight out of college, graduates from the country’s maritime academies can earn more than $200,000 as a commercial sailor, with free food and private accommodations... Despite the pay and perks, maritime jobs go begging, and it is raising national-security concerns.” Other selling points include “six months vacation, live wherever you want, and you’re serving the nation” and onboard “gyms, connectivity, and cuisine”. The catch is that you have to be at sea for months at a time. 42: Study (h/t @KierkegaardEmil): there was minimal “learning loss” from COVID school closures, best estimate is “0.02 standard deviations per 100 days of school closure”. I correctly predicted this back in 2021, but I also wrote in March of this year about how there’s been a general decline in NAEP scores since then. It seems like maybe a student having their specific school closed for longer than other schools didn’t hurt them, but some sort of general cultural change, maybe related to COVID, did hurt. 43: Sam Bankman-Fried’s mother on why she thinks his trial was unfair. SBF is appealing his conviction and will probably be making some of these same points in court. Can’t find a prediction market directly on the appeal, but this one says only 15% chance he serves under 10 years, this one says 15% chance of a Trump pardon, so it doesn’t seem like there’s much room for him to be freed (or get a significantly shorter sentence) on appeal. And Wired says that only 5-10% of appeals like these succeed. 44: Related: Trump pardons Juan Orlando Hernandez, former Honduran president extradited to the US for narco-corruption. Some sources are trying to find a Prospera angle - Prospera and other ZEDEs were approved under JOH’s administration, and the Prosperans seem to have good MAGAworld connections - but I don’t think this is their top priority, and I don’t know if it requires much explanation for Trump to be pro-right-wing Latin American politicians convicted by the Biden administration. More interesting is that apparently JOH and SBF were cellmates (X), “SBF spent extensive time helping JOH with trial prep” and SBF told an interviewer that “Juan Orlando is the most innocent prisoner I’ve met, myself included.” ChatGPT is not impressed with the Trump/SBF case for JOH’s innocence. Related: JOH’s conservative party on track to win this month’s extremely-close Honduran elections, great news for Prospera if it happens. 45: The “100 Above The Park” building in St Louis (h/t Bobby Fijan on X): 46: The death toll of the ongoing Sudan genocide has risen to about 150,000. Nicholas Kristof writes that the world has once again failed to prevent atrocities, and argues that the most important point of leverage is pressure on the United Arab Emirates, which is arming the genociders. Sam Kriss also writes about the situation in The World’s First Matcha Labubu Genocide, but is unimpressed with Kristof’s take: Sudan is passed over in a deeply uncomfortable silence. The absolute most you can do is blame the Emiratis. From what I’ve seen, more people seem to be appalled at the UAE for its frankly marginal role in arming the RSF than at the RSF itself. This is the approved way of understanding any inscrutably indigenous foreign conflict: you just worm out any third-party involvement and then act like you’ve solved the whole thing. I side with Kristof here, for reasons that Sam himself touches on later in his piece, in a section comparing Darfur with Gaza. It would be very easy to make people care about Darfur again. All it would take is a loud, vocal contingent of RSF apologists in the Western media. I agree, but would frame it less cynically: the reason Westerners pay attention to Gaza is that there’s a lever to push: not only does America support Israel, but many of their friends support Israel, so they can imagine convincing America or at least their friends to stop, and at least feel like there is some remote chance of making a small difference (and in fact, Trump getting mad at Israel and deciding to pressure them was decisive in effecting the cease-fire). On the other hand, we don’t have many levers to affect ethnic Baggara in the Rapid Support Forces of Sudan, so it doesn’t really feel useful to write blog posts arguing that they should stop; obviously they should stop, nobody disagrees with this, and it goes without saying - so nobody says it. But the US does support the UAE, and many of our friends like the UAE or at least go there on vacation, so maybe it’s possible to have make some small difference by embarrassing them. 4D chess take is that Sam Kriss agrees with all of this, but “loudly” and “vocally” argued against it to give people like me a hook to write about this genocide with, in which case I thank him for his sacrifice. It would also be nice to be able to donate, but I don’t know who to trust in the region - other than Doctors Without Borders, who are usually pretty good. 47: The AI Futures Project (group of AI-will-be-fast intellectuals) and the AI As A Normal Technology team (group of AI-will-be-slow intellectuals) wrote an adversarial collaboration in Asterisk explaining what they agree on, for example: That there’s an important distinction between existing AI and “strong AGI”

Astral Codex Ten

Table of Contents

Atlas

Herasight

Herasight

Article

Metadata

Appears In

External Links

Source Context

Backlinks

Astral Codex Ten

Table of Contents

Atlas

Herasight

Herasight

Article

Metadata

Appears In

Related Pages

External Links

Source Context

Backlinks