YourNetworkIsHaunted

Thank you for providing some actual domain experience to ground my idle ramblings.

I wonder if part of the reason why so many high profile intellectuals in some of these fields are so prone to getting sniped by the confabulatron is an unwillingness to acknowledge (either publicly or in their own heart) that "random bullshit go" is actually a very useful strategy. It reminds me of the way that writers will talk about the value of just getting words on the page because it's easier to replace them with better words than to create perfection ex nihilo, or the rubber duck method of troubleshooting where just stepping through the problem out loud forces you to organize your thoughts in a way that can make the solution more readily apparent. It seems like at least some kinds of research are also this kind of process of analysis and iteration as much as if not more than raw creation and insight.

I have never met Donald Knuth, and don't mean to impugn his character here, even as I'm basically asking if he's too conceited to properly understand what an LLM is, but I think of how people talk about science and scientists and the way it gets romanticized (see also Iris Merideth's excellent piece on "warrior culture" in software development) and it just doesn't fit a field that can see meaningful progress from throwing shit at the wall to see what sticks. A lot of the discourse around art and artists is more willing to acknowledge this element of the creative process, and that might explain their greater ability and willingness to see the bullshit faucet for what it is. Maybe because science and engineering have a stricter and more objective pass/fail criteria (you can argue about code quality just as much as the quality of a painting, but unlike a painting either the program runs or it doesn't. Visual art doesn't generally have to worry about a BSOD) there isn't the same openness to acknowledge that the affirmative results you get from an LLM are still just random bullshit. I can imagine the argument being: "The things we're doing are very prestigious and require great intelligence and other things that offer prestige and cultural capital. If 'random bullshit go' is often a key part of the process then maybe it doesn't need as much intelligence and doesn't deserve as much prestige. Therefore if this new tool can be at all useful in supplementing or replicating part of our process it must be using intelligence and maybe it deserves some of the same prestige that we have."

He is altering the deal. Pray he does not alter it further. These are definitely the good guys, right?

Even in Knuth's account it sounds like the LLM contribution was less in solving the problem and more in throwing out random BS that looked vaguely like different techniques were being applied until it spat out something that Knuth and his collaborator were able to recognize as a promising avenue for actual work.

His bud Filip Stappers rolled in to help solve an open digraph problem Knuth was working on. Stappers fed the decomposition problem to Claude Opus 4.6 cold. Claude ran 31 explorations over about an hour: brute force (too slow), serpentine patterns, fiber decompositions, simulated annealing. At exploration 25 it told itself “SA can find solutions but cannot give a general construction. Need pure math.” At exploration 30 it noticed a structural pattern in an earlier solution. Exploration 31 produced a working construction.

I am not a mathematician or computer scientist and so will not claim to know exactly what this is describing and how it compares to the normal process for investigating this kind of problem. However, the fact that it produced 4 approaches over 31 attempts seems more consistent with randomly throwing out something that looks like a solution rather than actually thinking through the process of each one. In a creative exploration like this where you expect most approaches to be dead ends rather than produce a working structure maybe the LLM is providing something valuable by generating vaguely work-shaped outputs that can inspire an actual mind to create the actual answer.

Filip had to restart the session after random errors, had to keep reminding Claude to document its progress. The solution only covers one type of solution, when Claude tried to continue another way, it “seemed to get stuck” and eventually couldn’t run its own programs correctly.

The idea that it's ultimately spitting out random answer-shaped nonsense also follows from the amount of babysitting that was required from Filip to keep it actually producing anything useful. I don't doubt that it's more efficient than I would be at producing random sequences of work-shaped slop and redirecting or retrying in response to a new "please actually do this" prompt, but of the two of us only one is demonstrating actual intelligence and moving towards being able to work independently. Compared to an undergrad or myself I don't doubt that Claude has a faster iteration time for each of those attempts, but that's not even in the same zip code as actually thinking through the problem, and if anything serves as a strong counterexample to the doomer critihype about the expanding capabilities of these systems. This kind of high-level academic work may be a case where this kind of random slop is actually useful, but that's an incredibly niche area and does not do nearly as much as Knuth seems to think it does in terms of justifying the incredible cost of these systems. If anything the narrative that "AI solved the problem" is giving Anthropic credit for the work that Knuth and Stapprrs were putting into actually sifting through the stream of slop identifying anything useful. Maybe babysitting the slop sluice is more satisfying or faster than going down every blind alley on your own, but you're still the one sitting in the river with a pan, and pretending the river is somehow pulling the gold out of itself is just damn foolish.

I mean, I can understand the argument that Anthropic at least maintained a fig leaf of ethics, but notably based on Saltman's statements OpenAI does still feel the obligation to maintain those optics, they're just not nearly as credible at doing so.

I actually dug up the context to make sure I wasn't forgetting something horrific. It's from a 2017 piece (CW: SSC Link) back before he went mask-off but was firmly in the "I'm a liberal and I talk exclusively about how liberals and their institutions suck" useful idiot phase of his career, so the overall essay is about how actually the wing nuts have a point when they say that all so-called neutral institutions are actually secret communist indoctrinators that want to trans your children and take your guns. I'm paraphrasing, obviously; he believes/pretends that when they called these things left-wing they didn't mean "literally in league with Stalin and the Devil". However, in the middle of the usual beigeness he tries to maintain his air of neutrality by having a section on how bad Voat ended up being, which concludes with:

The moral of the story is: if you’re against witch-hunts, and you promise to found your own little utopian community where witch-hunts will never happen, your new society will end up consisting of approximately three principled civil libertarians and seven zillion witches. It will be a terrible place to live even if witch-hunts are genuinely wrong.

God that was bleak - I thought Nick was bad in his guest spots on Alex's show (seen via Knowledge Fight, of course) but apparently you really do need at least two layers of insulating podcast to avoid suffering critical psychic damage from that level of hatred. I appreciated the acknowledgement that in order to feel at all okay playing clips you needed to sanewash him a little bit. I'm pretty sure that JorDan do the same thing with Alex and don't acknowledge it nearly often enough.

I also feel like some of Nick's schtick is about trying to position himself and maintain his position in the right wing grifter bigot-industrial complex. Like, the open disdain for his audience and presenting his actually pretty straightforward feelings on the halftime show as somehow brave and iconoclastic is also about differentiating himself and making his audience feel superior to Alex, Tucker, Candace, etc. In that sense the open disdain for the audience serves another purpose in terms of reinforcing heirarchy. Look at how great it feels for me to be better than you. And even you are better than the chuds, who are better than the racialized other.

It's especially strange because becoming less prone to bias and developing a clear understanding of what serves your interest is so much of the pitch for Rationalism as a community/ideology/project. Like, here's unbearably long essays that promise to help cultivate the superpower of seeing the world clearly and acting in it effectively, now if you acknowledge that nobody outside this small set of group homes is actually doing that you'll be shunned. And that's not getting into how easily exploitable those assumptions of good faith are by bad-faith actors. It comes back to that quote from Scott that has stuck in my head apparently more than it did his: if you build a community based on the principle that you will absolutely never have a witch hunt you will end up living among approximately seven principles civil libertarians and eleven million goddamn witches, and this is true even if you're right that witch hunts are bad.

That is what happens when your mode of analysis is closer to erotic Harry Potter fan fiction (which is indeed the medium in which Yudkowsky has delivered some of his prognostications)

I was going to throw a point of order about not all fanfic being erotic, but given how they fetishize "intelligence" and "rationality" I can't be sure that they don't get off on that slog.

There's gotta be a pithy way of talking about this. I propose "Phantom Funds" - money that investors and analysts expect to be there that ends up not existing when the cards are turned over.

Why yes this does largely boil down to fraud but without the legal consequences.

Is it though? Like, theres missing detail about the request to publish their talking points in addition to the original reporting, but it's a pretty fair description of the original reporting at issue. That's pretty solid as far as headlines go.

It's such a powerful dodge. What you're actually saying is "we're going to keep doing exactly what we're doing and see if that fixes it" because the nature of innovation is such that it's actually pretty complex to "invest" in, and very rarely has the direct application you need. Like, you don't get penicillin by investing in pharmaceutical innovation you get it by paying some nerd to fuck off to the jungle for a few years and hope that his special interest ends up being useful. Bell Labs was able to basically invent the modern world by funneling the profits of their massive monopolistic empire into ~~a bunch of nerds poking stuff with probes to see what happens~~ elementary physics and materials science research that didn't have a definite objective.

I've heard worse ideas. It's funny; I would have expected the people who were in tech because it looked like the best bet for a relatively stable in-demand career would have been the ones who were crap at it relative to the folks doing it purely for love of the game. But it turns out that having something else going on is closely linked to touching grass in ways that make you harder to lure into the cult.