Benquo

Karma: 13,091

Benquo 9 Apr 2026 13:08 UTC
2 points
0
in reply to: cdt’s comment on: We can prevent progress! Conceptual clarity, and inspiration from the FDA
Progressivism seems like a third degree simulacral procession from Whig Theory of History, which is at least a theory. Whig Theory of History relates to progress as a contingent claim to be asserted about aggregates, which is meaningful because it lawfully decomposes into constituent facts we care about. Progressivism seems to inherit the emotional loading of the term as a given and then situate intentions relative to it.

Progress Studies and Works in Progress seem to be relying on “progress” as something whose value & reality is uncontested even if its specific meaning is unclear and the rollout is uneven.

Benquo 9 Apr 2026 12:55 UTC
2 points
0
on: Beliefs are Chosen to Serve Goals
Am I missing something or does this piece never actually make the argument it promises? It seems roughly comparable to:

A Turing machine & its starting tape are a joint specification; neither means or can do anything without the other, and the same functions can be differently allocated between them.

I wouldn’t suggest this breaks the orthogonality of Turing machines. Nor does the fact that a small machine with a small tape can’t compute large functions, and a small enough machine isn’t Turing complete, break the relevant orthogonality claim.

Similarly, I don’t understand why you think that beliefs and values being only jointly predictive of actions (and therefore jointly selected) restricts the space of values or propositional beliefs implementable in an AI within the space of expressible ones.

Benquo 8 Apr 2026 16:45 UTC
2 points
−3
in reply to: Mateusz Bagiński’s comment on: We can prevent progress! Conceptual clarity, and inspiration from the FDA
Those seem like attempts to extend the useful life of the current regime by trying to organize around doing more of the things that originally won it legitimacy, rather than to productively criticize or supersede it. Sometimes you should patch up an old thing rather than buy a new one, sometimes this is false economy because the cost of upkeep is higher than the amortized cost of replacement, and sometimes you’re driving around in an explosive death trap or breathing mold every day making you sick when you should really just get a safe new car or house built from scratch.

I would put Tyler Cowen in the same category, accepting things like GDP as the best politically available target to organize around, but trying to persuade people to do good rather than bad things to raise the GDP.

Benquo 8 Apr 2026 14:43 UTC
23 points
4
on: An easy coordination problem?
Some complicating factors that might help explain:

Owned vs borrowed power https://medium.com/@samo.burja/borrowed-versus-owned-power-a8334fbad1cd

Altman, Amodei, and Hassabis have borrowed power that may be conditioned on their AI progress. Musk’s situation is more complex. He has more notional ownership, which means he has more short-run control, but on some timeframes he’s still dependent on access to capital that depends on growth expectations, though these are much less specific to AI. Xi’s power is probably least entangled with this specific thing; Thiel’s suggested that the Chinese interest in AI is mostly about internal mass surveillance, though accountability and transparency to the executive more generally seems to me like a better fit for Xi’s problems: https://benjaminrosshoffman.com/doge-in-context/

These people are crazy.

Many of these people been strongly selected and conditioned for maniacal dedication to progress, so asking them to notice that that’s not in their interest is a difficult: https://benjaminrosshoffman.com/approval-extraction-advertised-as-production/

Their “belief” in ASI is cynical opportunism or otherwise deeply confused.

Overpromising and then pivoting is normal for startupworld. If the story is that what you’re working on is world-destroying-level dangerous, and that’s picked up in the vibe, investors just hear that you’re doing something powerful and transgressive, kind of like Uber. And no one’s really worried that Uber will kill everyone. Musk’s concerns about AI are notoriously confused: https://benjaminrosshoffman.com/openai-makes-humanity-less-safe/

That also explains why “cut a deal with Sam Altman” is not an appealing option to Musk; he already did that!

Benquo 8 Apr 2026 14:42 UTC
6 points
−2
on: We can prevent progress! Conceptual clarity, and inspiration from the FDA
Our culture is dominated by an ideology of progress, called progressivism, which conflates purportedly inevitable progressions along trendlines (especially ones that amount to an increase in taxable activity, or increased economic mobilization) with the solution of problems and people’s lives getting better. Because it’s an ideology, progressives worship progress rather than having honest propositional views on it amenable to evidence; enough evidence might cause a personal crisis, but they don’t have the virtues of lightness or specificity about it.

In practice progressivism is a job creation scheme for elites. The jobs created have to be constituted to manage problems, rather than solve them; solving problems destroys jobs. Oddly, if you wanted to slow down the rate at which we solve our proximate problems using AI or increase AI capacities, the best option available short of revolution or radical dissidence en masse might be to make “AI Progress” an important positive policy goal and try to persuade elites that it’s important to get those metrics up.

Institutional recommendations are shaped by implicit constraints like “don’t reduce headcount” and “don’t invalidate your department’s premise,” internalized as limits on what’s thinkable: https://benjaminrosshoffman.com/parkinsons-law-ideology-statistics/#diagnosis

Calling for X produces jobs doing the opposite: https://benjaminrosshoffman.com/openai-makes-humanity-less-safe/

Neoclassical (progressive) economics tries to quantify “total value,” and then maximize it, which in practice means maximizing transaction volume: https://substack.com/@benhoffman700141/note/c-237461608

A much deeper dive into the same thing: https://benjaminrosshoffman.com/the-domestic-product/

Benquo 8 Apr 2026 4:02 UTC
3 points
1
in reply to: NoSignalNoNoise’s comment on: Telescopes Need Good Lenses
I’d think that if their American analogues didn’t also reproduce way above replacement.

Telescopes Need Good Lenses

Benquo8 Apr 2026 1:25 UTC

18 points

2 comments11 min readLW link

(benjaminrosshoffman.com)

Benquo 7 Apr 2026 16:10 UTC
2 points
0
in reply to: xpym’s comment on: The Corner-Stone
Calvinist England is a well documented culturally near case.

China’s track record suggests a more recent positive trajectory, so that their measured quality of life seems on par with the US despite the US’s vastly superior per capita endowment as recently as 1980, and their industrial and technological capacity is not unambiguously inferior.

Benquo 5 Apr 2026 13:41 UTC
2 points
0
in reply to: Wei Dai’s comment on: The Corner-Stone
As I understand your Smart Losers toy model, there’s an iterated prisoner’s dilemma, and on some rounds knowable only to “smart” players pairings and results are not published afterwards. There’s no other functional meaning of “smart” here, non-adversarial uses of intelligence don’t exist. So non-”smart” players start with grim trigger, but “smart” players grim trigger in published rounds and defect in unpublished rounds, which can push everyone into a defect-defect equilibrium.

It should be obvious from my characterization what I think the main weakness in this model is. I wrote Civil Law and Political Drama to try to characterize the difference and relation between adversarial and nonadversarial consciousness more precisely.

But that’s something of a timeless description, and doesn’t explain why one or the other would win at some particular time. An important contingent factor is the balance between external performance pressure, and the opportunity for central allocation of patronage. Systems with a large stable extracted surplus like tax revenue select, in the absence of specific countermeasures, for courtiers seeking patronage. Without specific countermeasures, this blows up catastrophically. The countermeasure that seems to work is an anti-norm: patronage-seekers only manage problems for the short-term benefit of the class as a whole, rather than either solving them or increasing them unboundedly to maximize their own share of the rents. I think rent-seeking classes that manage this coordinate through perverted shame.

So we see functionally oriented cognition more in cases where groups are under strong external performance pressure, or in the rare punctuating cases where pronormative actors developed enough awareness of themselves as a class to take collective action against their enemies. (Norman Gottwald’s The Tribes of Yahweh argues that the ancient Israelites did something similar, though I’m still making sense of that situation, and it’s harder to evaluate because records are much worse.)

This isn’t a complete model, and it would be helpful for others to check it against alternative explanations, check it for coherence, and develop the connections between different parts.

Much of this overlaps strongly with James Camacho’s comment thread, I think it would be better for the three of us to be discussing this jointly rather than separately; I’d like to see where you find either of us unpersuasive and why.

Benquo 5 Apr 2026 12:51 UTC
13 points
0
in reply to: alseph’s comment on: The Corner-Stone

It feels like all employers now want candidates who are obsessed with the role they applied for.”

Or they want candidates willing to role-play that obsession. This seems much more probable on priors, easier to access from any plausible naïve view of the world if you’re not unseeing adversariality, and at least as consistent with the data.

I implore you to consider that different times call for radically different sorts of work, and you now have sufficient evidence against the hypothesis that the right thing to do is try to contribute to the shared project our governing institutions are recruiting you for. John Adams formulated this very tidily:

The Science of Government it is my Duty to study, more than all other Sciences: the Art of Legislation and Administration and Negotiation, ought to take Place, indeed to exclude in a manner all other Arts. I must study Politicks and War that my sons may have liberty to study Mathematicks and Philosophy. My sons ought to study Mathematicks and Philosophy, Geography, natural History, Naval Architecture, navigation, Commerce and Agriculture, in order to give their Children a right to study Painting, Poetry, Musick, Architecture, Statuary, Tapestry and Porcelaine.

If you find yourself confronting not rule of law, strong property rights, and voluntary mutual exchange, but something more like Hobbes’s state of nature:

In such condition there is no place for industry, because the fruit thereof is uncertain: and consequently no culture of the earth; no navigation, nor use of the commodities that may be imported by sea; no commodious building; no instruments of moving and removing such things as require much force; no knowledge of the face of the earth; no account of time; no arts; no letters; no society; and which is worst of all, continual fear, and danger of violent death; and the life of man, solitary, poor, nasty, brutish, and short.

then you’re in the first stage of Adams’s developmental pyramid, and work on the later stages is only helpful in special cases, not generically.

Cf also Ecclesiastes 3:

To every thing there is a season, and a time to every purpose under the heaven:
A time to be born, and a time to die; a time to plant, and a time to pluck up that which is planted;
A time to kill, and a time to heal; a time to break down, and a time to build up;
A time to weep, and a time to laugh; a time to mourn, and a time to dance;
A time to cast away stones, and a time to gather stones together; a time to embrace, and a time to refrain from embracing;
A time to get, and a time to lose; a time to keep, and a time to cast away;
A time to rend, and a time to sew; a time to keep silence, and a time to speak;
A time to love, and a time to hate; a time of war, and a time of peace.
What profit hath he that worketh in that wherein he laboureth?

Benquo 5 Apr 2026 12:32 UTC
2 points
0
in reply to: DanB’s comment on: The Corner-Stone
“than we can effectively use” is not an exogenous quantity. More people means more available intelligence, but also means more users of intelligence! So, yes, it can be “that bad” if the counterfactual where our promotion mechanisms weren’t fanatically devoted to crippling the best of us is sufficiently better than our actual reality.

Benquo 5 Apr 2026 0:45 UTC
6 points
0
in reply to: Benquo’s comment on: Chicken-Free Egg Whites
If you really want to go crazy, here are two previously impossible things you can try:
1. A near-dry protein foam folded into melted chocolate before it sets could give you something like aerated chocolate with a different cell structure than what you get from pressurized gas injection.
2. Twice-cooked meringue pops: Put something on a stick (could be something like a preformed cake pop) and coat it with a layer of ovalbumin meringue (maybe 1 part ovalbumin, 3 parts sugar, 2 parts water by weight, plus a little of whatever flavor-bearing addition you like). Mount the stick vertically on an oven-safe container to bake it until the meringue sets. Once it sets and cools, deep fry it. This might allow a thick but light, airy, and crunchy shell. You could do a small baked Alaska pop this way, refreeze it, and then deep fry it.

Benquo 4 Apr 2026 22:48 UTC
6 points
0
on: Chicken-Free Egg Whites
Macarons are the obvious target, since people already age whites partly to shift water balance, and powder gives you that dial directly. But macarons are sensitive to everything, so also the highest-risk experiment.

The biggest gain for meringue specifically is probably flavor loading: more protein network per unit of cocoa or fruit powder means the foam can carry more before it collapses. They might also bake faster and more reliably; lower hydration should help with tackiness, hollow shells, and humidity sensitivity.

Other recipes where egg whites are currently the limiting water source include flourless chocolate cakes, nut tortes, high-fat foams, and mousses. Concentrated whites could let you add aeration without thinning the base.

Benquo 4 Apr 2026 21:16 UTC
2 points
0
in reply to: James Camacho’s comment on: The Corner-Stone
Very pleasing to watch you work through the relevant theory much faster than I did (albeit also with more directly relevant reading materials).

Bounded cognition is relevant here. Negotiating coalitional status competes for cognitive resources with participating in shared models of external reality, so if you enforce norms well enough, it’s cheaper for people to mostly just be normative.

The English Civil War is a historical example of a pronormative revolution against an antinormative regime; I wrote about it in Calvinism as a Theory of Recovered High-Trust Agency (also linked in a footnote to the post, but I don’t expect anyone followed all the links). I think it’s an important case study; the major benefit of Calvinism seems to be having any sort of idea of the existence of a conflict between pronormative and antinormative mindsets, while endorsing the normative side.

The major disadvantage otherwise pronormative cognition has is naïvete to the existence of antinormativity, sometimes to the point of the sort of unseeing I complained about in the Lacanian section of Compradorization. Civil Law and Political Drama is my best attempt to lay out single framework that can describe both pronormative and antinormative cognition.

Benquo 4 Apr 2026 4:13 UTC
7 points
0
in reply to: James Camacho’s comment on: The Corner-Stone
I think I need to clarify something more abstract here:

There’s optimization against norms as such, what Jessica Taylor called anti-normativity. Clarifying what basic normativity implies can seem to produce easy object-level wins relative to baseline, but not because it’s intrinsically difficult. (Much of the Sequences is about what it looks like to be wrong on purpose and what it would look like not to.) And until you know about antinormativity, not only is the need for clarification itself a puzzle in need of an answer, but all your object-level gains are vulnerable to expropriative ambush by an adversary you’re unaware of.

Benquo 4 Apr 2026 2:37 UTC
3 points
0
in reply to: James Camacho’s comment on: The Corner-Stone
I think the idea that one needs “training in game theory and information theory” to try to recruit usable talent rather than trying to do what everyone else is doing, is the kind of backwardsness the current educational system instills by design. One does need some willingness to believe that everyone else might be wrong, of course, but then one can just make reasonable guesses and use trial and error. When people are trying to do things wrong on purpose, the problem is not that they lack specialized technical knowledge in means-ends reasoning, but that they’re trying to signal loyalty instead of solving the problem they’re pretending to try to solve. The difficult problem that requires theory is how not to become inappropriately arrogant in the process, by mistaking everyone else’s perversity for your brilliance.

The idea that education is mostly unrelated to the systems staffed almost entirely by people who’ve spent over a decade and often almost two being conditioned by the education system and saying “please may I have some more,” and to the systems that fund the education system and define and enforce its special privileges, seems so totally implausible to me that I don’t even know what to say in response, aside from this paragraph.

Marxism is not the origin of the assault on education; American Progressives were overtly attacking literacy and cognition through the educational system long before Marxism became popular in academia.

Benquo 3 Apr 2026 23:15 UTC
2 points
0
in reply to: James Camacho’s comment on: The Corner-Stone
There is a central policy of creating patronage jobs. I wrote about the causes of this in The Debtors’ Revolt, what it looks like in practice in Parkinson’s Law and the Ideology of Statistics, and historical precedent for patronage in What is a republic? A Roman aristocratic perspective.

Schools create jobs for teachers by stretching basic education over an improbable number of years. Professional-Managerial Class apparatchiks (I justify this abstraction in the linked posts) understand their main responsibility to be creating more jobs for people like themselves, which creates runaway intermediation like we see in recruiters and HR, instead of easy active strategies like sending letters to all National Merit Semifinalists or SAT high scorers, walking into a well-run McDonald’s and trying to poach the manager, etc. Finance is heavily regulated and functions as a mechanism for preferentially allocating access to cheap credit to the sorts of people who play the game.

Benquo 3 Apr 2026 17:42 UTC
3 points
0
in reply to: Eliezer Yudkowsky’s comment on: Is fever a symptom of glycine deficiency?
On linoleic acid as a “dietary villain,” I’m familiar with two plausible stories:

1 PUFAs generally as oxidative liability (via Masterjohn), which seems mechanistically tight but doesn’t seem to generate a clear observational signal so the effect size is probably small, though I can think of ways to save the hypothesis (snapshots of PUFA consumption may not reflect AUC or PUFA levels). This should only be a problem if you’re storing rather than burning PUFAs, so I’d expect a lot of heterogeneity in response depending on many other factors. It’s also not specific to LA except insofar as LA is a common PUFA in Western diets.

2 Brad Marshall’s torpor story. I looked into this a bit, and while there aren’t experimental results on humans, there are both positive and negative results in animals; the main negative one I found was Gerson et al. (2008). I’m skeptical of external validity as applied to a species that doesn’t have what seem to be the relevant physiological torpor mechanisms.

Am I missing something big here?

Benquo 3 Apr 2026 16:38 UTC
2 points
0
in reply to: ChristianKl’s comment on: Your Fascia Doesn’t Recognize You as a Hunter
Blocking endogenous HA receptor binding in mice impairs gut development and barrier function (Riehl et al. 2015) within about 5 days, so we know that signal is generated routinely. Immunostaining of healthy human jejunum also shows intense HA in the lamina propria with no staining in the epithelial layer, so healthy guts do produce substantial HA endogenously. That doesn’t tell us whether it’s produced steadily or mainly during episodic damage and repair, or whether epithelial HA produces different effects, or whether supplement levels send an important extra signal.

The Bacteroides as pathogen shield argument is new and interesting. I’d expect hyaluronidase producing bacteria to roughly equilibrate to available HA over time, whether from endogenous production, episodic repair, or diet. Do you have specific reasons to think supplementing HA would change the ratio rather than just the levels?

Benquo 3 Apr 2026 3:20 UTC
4 points
2
in reply to: Eliezer Yudkowsky’s comment on: Is fever a symptom of glycine deficiency?
If you eat 3 oz of plain pork rinds, you get 52g of protein and 26g of fat. Of that protein, 19% is glycine, 10g. Of that fat, 11% is linoleic acid, 2.9g, which is 16% of average American daily intake. Of course, you might be unusually adversely sensitive to linoleic acid, which would be a reason to get glycine some other way.

Chicken is a much less efficient glycine delivery vehicle. If you eat a pound of roasted skin-on chicken (roughly half a Costco rotisserie chicken) you get 125g of protein and 62g of fat. Of that protein, 6.5% is glycine, 8g. Of that fat, 19% is linoleic acid, 12g, which is ²⁄₃ of the average American daily intake.

Chicken skin alone is even worse. If you eat ¹⁄₃ kg of roasted chicken skin (wow, that’s a lot), you get 67g of protein and 137g of fat. Of that protein, approximately 13% is glycine, 9g. Of that fat, 20% is linoleic acid, 27g, which is 1.5x the average American daily intake.

I didn’t know that until I looked up the numbers just now. It probably varies a lot based on what the animals are fed. Typical Chinese chickens are probably meaningfully better than the American ones on this metric, as the Chinese tend to prefer chickens that taste like chicken over ones bred to maximize meat yields above all else, though this is changing.

If you want to avoid linoleic acid, and want to get glycine from whole foods, best to do things like eat tendons and skimmed bone broth.

Benquo

Te­lescopes Need Good Lenses

Telescopes Need Good Lenses