Stubsack: weekly thread for sneers not worth an entire post, week ending 28th December 2025 - awful.systems
awful.systems
Want to wade into the snowy surf of the abyss? Have a sneer percolating in your system but not enough time/energy to make a whole post about it? Go forth and be mid: Welcome to the Stubsack, your first port of call for learning fresh Awful youā€™ll near-instantly regret. Any awful.systems sub may be subsneered in this subthread, techtakes or no. If your sneer seems higher quality than you thought, feel free to cutā€™nā€™paste it into its own post ā€” thereā€™s no quota for posting and the bar really isnā€™t that high. > The post Xitter web has spawned soo many ā€œesotericā€ right wing freaks, but thereā€™s no appropriate sneer-space for them. Iā€™m talking redscare-ish, reality challenged ā€œculture criticsā€ who write about everything but understand nothing. Iā€™m talking about reply-guys who make the same 6 tweets about the same 3 subjects. Theyā€™re inescapable at this point, yet I donā€™t see them mocked (as much as they should be) > > Like, there was one dude a while back who insisted that women couldnā€™t be surgeons because they didnā€™t believe in the moon or in stars? I think each and every one of these guys is uniquely fucked up and if I canā€™t escape them, I would love to sneer at them. (Credit and/or blame to David Gerard for starting this. Merry Christmas, happy Hannukah, and happy holidays in general!)

Want to wade into the snowy surf of the abyss? Have a sneer percolating in your system but not enough time/energy to make a whole post about it? Go forth and be mid: Welcome to the Stubsack, your first port of call for learning fresh Awful youā€™ll near-instantly regret.

Any awful.systems sub may be subsneered in this subthread, techtakes or no.

If your sneer seems higher quality than you thought, feel free to cutā€™nā€™paste it into its own post ā€” thereā€™s no quota for posting and the bar really isnā€™t that high.

The post Xitter web has spawned soo many ā€œesotericā€ right wing freaks, but thereā€™s no appropriate sneer-space for them. Iā€™m talking redscare-ish, reality challenged ā€œculture criticsā€ who write about everything but understand nothing. Iā€™m talking about reply-guys who make the same 6 tweets about the same 3 subjects. Theyā€™re inescapable at this point, yet I donā€™t see them mocked (as much as they should be)

Like, there was one dude a while back who insisted that women couldnā€™t be surgeons because they didnā€™t believe in the moon or in stars? I think each and every one of these guys is uniquely fucked up and if I canā€™t escape them, I would love to sneer at them.

(Last substack for 2025 - may 2026 bring better tidings. Credit and/or blame to David Gerard for starting this.)

  • gerikson@awful.systemsEnglish
    16Ā·
    18 days ago

    A rival gang of ā€œAIā€ ā€œresearchersā€ dare to make fun of Big Yudā€™s latest book and the LW crowd are Not Happy

    Link to takedown: https://www.mechanize.work/blog/unfalsifiable-stories-of-doom/ (hearbreaking : the worst people you know made some good points)

    When we say Y&Sā€™s arguments are theological, we donā€™t just mean they sound religious. Nor are we using ā€œtheologicalā€ to simply mean ā€œwrongā€. For example, we would not call belief in a flat Earth theological. Thatā€™s because, although this belief is clearly false, it still stems from empirical observations (however misinterpreted).

    What we mean is that Y&Sā€™s methods resemble theology in both structure and approach. Their work is fundamentally untestable. They develop extensive theories about nonexistent, idealized, ultrapowerful beings. They support these theories with long chains of abstract reasoning rather than empirical observation. They rarely define their concepts precisely, opting to explain them through allegorical stories and metaphors whose meaning is ambiguous.

    Their arguments, moreover, are employed in service of an eschatological conclusion. They present a stark binary choice: either we achieve alignment or face total extinction. In their view, thereā€™s no room for partial solutions, or muddling through. The ordinary methods of dealing with technological safety, like continuous iteration and testing, are utterly unable to solve this challenge. There is a sharp line separating the ā€œbeforeā€ and ā€œafterā€: once superintelligent AI is created, our doom will be decided.

    LW announcement, check out the karma scores! https://www.lesswrong.com/posts/Bu3dhPxw6E8enRGMC/stephen-mcaleese-s-shortform?commentId=BkNBuHoLw5JXjftCP

    Update an LessWrong attempts to debunk the piece with inline comments here

    https://www.lesswrong.com/posts/i6sBAT4SPCJnBPKPJ/mechanize-work-s-essay-on-unfalsifiable-doom

    Leading to such hilarious howlers as

    Then solving alignment could be no easier than preventing the Germans from endorsing the Nazi ideology and commiting genocide.

    Ummm pretty sure engaging in a new world war and getting their country bombed to pieces was not on most Germanā€™s agenda. A small group of ideologues managed to sieze complete control of the state, and did their very best to prevent widespread knowledge of the Holocaust from getting out. At the same time they used the power of the state to ruthlessly supress any opposition.

    rejecting Yudkowsky-Soaresā€™ arguments would require that ultrapowerful beings are either theoretically impossible (which is highly unlikely)

    ohai begging the question

    • scruiser@awful.systemsEnglish
      10Ā·
      17 days ago

      A few commentsā€¦

      We want to engage with these critics, but there is no standard argument to respond to, no single text that unifies the AI safety community.

      Yeah, Eliezer had a solid decade and a half to develop a presence in academic literature. Nick Bostrom at least sort of tried to formalize some of the arguments but didnā€™t really succeed. I donā€™t think they could have succeeded, given how speculative their stuff is, but if they had, review papers could have tried to consolidate them and then people could actually respond to the arguments fully. (We all know how Eliezer loves to complain about people not responding to his full set of arguments.)

      Apart from a few brief mentions of real-world examples of LLMs acting unstable, like the case of Sydney Bing, the online appendix contains what seems to be the closest thing Y&S present to an empirical argument for their central thesis.

      But in fact, none of these lines of evidence support their theory. All of these behaviors are distinctly human, not alien.

      Even with the extent that Anthropicā€™s ā€œresearchā€ tends to be rigged scenarios acting as marketing hype without peer review or academic levels of quality, at the very least they (usually) involve actual AI systems that actually exist. It is pretty absurd the extent to which Eliezer has ignored everything about how LLMs actually work (or even hypothetically might work with major foundational developments) in favor of repeating the same scenario he came up with in the mid 2000s. Or even tried mathematical analyses of what classes of problems are computationally tractable to a smart enough entity and which remain computationally intractable (titotal has written some blog posts about this with material science, tldr, even if magic nanotech was possible, an AGI would need lots of experimentation and canā€™t just figure it out with simulations. Or the lesswrong post explaining how chaos theory and slight imperfections in measurement makes a game of pinball unpredictable past a few ricochets. )

      The lesswrong responses are stubborn as always.

      Thatā€™s because we arenā€™t in the superintelligent regime yet.

      Yā€™all arenā€™t beating the theology allegations.

      • blakestacey@awful.systemsEnglish
        8Ā·
        17 days ago

        Yeah, Eliezer had a solid decade and a half to develop a presence in academic literature. Nick Bostrom at least sort of tried to formalize some of the arguments but didnā€™t really succeed.

        (Guy in hot dog suit) ā€œWeā€™re all looking for the person who didnā€™t do this!ā€

    • corbin@awful.systemsEnglish
      7Ā·
      16 days ago

      I clicked through too much and ended up finding this. Congrats to jdp for getting onto my radar, I suppose. Are LLMs bad for humans? Maybe. Are LLMs secretly creating a (mind-)virus without telling humans? Thatā€™s a helluva question, you should share your drugs with me while we talk about it.