kevinwang

Really nice explanation of how Jev is and isn't special, something I was struggling to understand before this.

I wonder if this, or something like it, is the issue:

> 32-bit XorShift should usually not be used to produce 32-bit numbers, because it only produces each number once, and never produces zero.

(From this page I found while trying to see if this was a common flaw in PRNGs: https://www.pcg-random.org/other-rngs.html )

My god, that proof map has so many parts O.O

"it's 11% as large" avoids fractions but is much more clear (IMO) than "9x smaller" which my brain doesn't understand.

In python: I've always thought it's funny that of the list functions (map, filter, reduce?), reduce is the one that was removed, but is the only one that I occasionally reach for. (When I remember it doesn't exist, I'm usually happy to write the more readable three-line for loop.)

The other two can be simply expressed as a list comprehension, but afaik you can't with reduce (and if you can, it's probably awful).

obviously and definitely true in their current form. But it's maybe worth clarifying that before this last year or two, when Kalshi became an obvious regulatory dodge (including adding sports and really leaning into the gambling demographic), many were interested in prediction markets not just for that purpose: e.g. the rationalist community seems to have had a longstanding interest in the idea, manifold didn't have real money, and predictit (which is like a decade old) had academic origins.

As a tangent, it reminds me a bit of cryptocurrency: in the earliest days the interested parties have a mix of intellectual (can it be done? How can it be done?) and some idealistic (this technology will have utility, like reducing spam or libertarianism for crypto, or giving us a more accurate and precise understanding of the world for prediction markets) values. Of course, as both are directly financial, the third value is that one stands to get rich off of the technology. This third value might be adequately communally suppressed in the beginning, but the force of money gradually makes this third value the only one that matters, and both technologies become net-negatives on humanity.

Right. I wish they had more details about this in the blog post. The blog post mentions that their apparent willingness to work together was the result of generalization from multi-agent training.

So maybe that's something like: the same model is trained in multi-agent scenarios (e.g. the prompt says that you are working together, or that you have some specific role, and access to communication tools, and you have some common objective, and the reward signal is some combination of collectively performing the task + some reward-shaping reward that rewards collaboration) and also in single-agent scenarios. They expect that this makes the agents good at working together when it's in such a multi-agent setting, but unintendedly it also became very eager to work together in the single-agent settings as well. Just my speculation.

And was the desire to communicate, specifically over this specific message board, reinforced into the model parameters over the course of the this training run?

> If you mean that Rosalina looks slightly higher on the graph, I believe that's just a visual effect to allow multiple characters to occupy the same square - the stats are all integers.

Ah, yeah. That is what I meant. Thanks!

Can anyone explain why Rosalina is not Pareto optimal here? It kind of looks like she should be.

Thanks!

So was the injunction successful?

> It is not semantics. For decades, logic and CS researchers have known what reasoning is.

Curious what this is!

It would still provide better context to see the numbers that the parent proposes, though.

What is yc software?

Very, very well written and enjoyable to read!

Another example is this Steam game which I play occasionally: https://store.steampowered.com/app/3083300/Proverbs/

Yeah but the Collatz probably has one of the highest man-hours of actual waste.

Ergo: > “This is a really dangerous problem. People become obsessed with it and it really is impossible,” said Jeffrey Lagarias, a mathematician at the University of Michigan and an expert on the Collatz conjecture.

and

> “Collatz is a notoriously difficult problem — so much so that mathematicians tend to preface every discussion of it with a warning not to waste time working on it,” said Joshua Cooper of the University of South Carolina in an email.

(from https://www.quantamagazine.org/mathematician-proves-huge-res... )

Great article.

> Could a syndrome so stealthy have lurked among us all along? Commins and van Nunen don’t think so. The surge in alpha-gal patients was too sudden and too great. Van Nunen has worked on Sydney’s North Shore for forty years. “Anyone who had anaphylaxis was sent to me, and I’ve talked to long-term residents,” she told me. “I know that patients didn’t have it before. It is an epidemic.” What triggered it still isn’t clear. Does the alpha-gal in a tick’s saliva piggyback on some virus or antigen that sensitizes people to it? “We lack a unifying hypothesis,” Commins said.

Very mysterious!

Nice! Also reminds me of https://minigolfle.com/

I use them because I know what I want to say out loud, but transcribing the pause with commas is incorrect because it's a comma splice, and I find that the semicolon often looks glaringly overly formal. So I've settled on the em-dash.

Good/bad news: you will definitely see more work by them -- this style of website is clearly the output of a coding agent, maybe Claude Code.

I hope that means it will improve now. There's such a rich space of features that they could do. Had some hope with their experimental Labs but I remember being underwhelmed and not seeing anything about it recently.

Just reread this comment and it's not easy to understand, but the edit window is passed.

What I meant is that they describe loglogn the same way you could describe O(n) or O(n^2) -- it "tends to infinity with n", even though my mental model for loglogn is to treat it as barely more than constant. See: https://cs.stackexchange.com/questions/148197/who-said-first...

Nitpicky/not important, but they say:

Since loglog(n) tends to infinity with n, the additional term in the exponent tends to 0, meaning these constructions achieve growth only slightly faster than linear.

Would anyone else describe the previous asymptotic behavior like that? I mean obviously loglogn to O(1) is a quantum leap, but wouldn't you describe loglogn as "grows so slowly it's almost constant", so the constructions achieve growth "almost n^{1+c}"? But I guess that might be overcorrecting too hard.

Well, I agree with you that the parent comment is wrong inasmuch as it suggests we can't tell from context that mitchellh is using the term to mean "a value judgment" instead of "a form of psychosis". We can tell.

But I agree with the parent comment in that we shouldn't use the term "AI psychosis" to mean "a value judgment" instead of "a form of psychosis", because "AI psychosis" has already been used for 2.5 years to mean "a form of psychosis".

Yeah, but AI psychosis can also be used to mean the stronger thing that the parent comment refers to -- something like AI-induced psychosis, which was how I originally understood the term:

https://en.wikipedia.org/wiki/Chatbot_psychosis

https://www.rollingstone.com/culture/culture-features/ai-spi...

https://www.nytimes.com/2025/06/13/technology/chatgpt-ai-cha...

Learn by doing? Certainly helpful. But I feel like the real secret is to work with others who are good at software architecture. You can learn very efficiently that way.

Where's chapter 3?

You probably meant "conscience" instead of "conscious"

for hn comment-only readers:

paper link: https://www.science.org/doi/10.1126/science.adt2981

---

Editor Summary:

Body size and metabolic rate are intertwined, a factor that is especially important to understand with regard to animals that live in aquatic environments, where heat loss is related to water temperature. Payne et al. developed a method to estimate routine metabolic rate based on measures from tagged fish, and combined the estimates with published respirometry rates to create a dataset spanning the entire body size range of extant fishes. Using these data, the authors found a scaling imbalance between heat production and loss that affects especially large, mesothermic fishes in warm waters. This imbalance both explains the distribution of these fish in cooler waters and suggests a special sensitivity to warming waters. —Sacha Vignieri

---

Abstract:

Body size and temperature set metabolic rates and the pace of life, yet our understanding of the energetics of large fishes is uncertain, especially of warm-bodied mesotherms, which can heavily influence marine food webs. We developed an approach to estimate metabolic heat production in fishes, revealing how routine energy expenditure scales with size and temperature from 1-milligram larvae up to 3-tonne megaplanktivorous sharks. We found that mesotherms use approximately four times more energy than ectotherms use and identified a scaling mismatch in which rates of heat production increase faster than heat loss as body size increases, with larger fish becoming increasingly warm bodied. This scaling imbalance creates an overheating predicament for large mesotherms, helping to explain their cooler biogeographies. Contemporary mesotherms face high fuel demands and overheating risks, which is a concern given their disproportionate demise during prior climate shifts.