Behind the Bastards just wrapped their four-part series on the Zizzians, which has been a fun trip. Nothing like seeing the live reactions of someone who hasn’t been at least a little bit plugged into the whole space for years.

I haven’t finished part 4, but so far I’ve deeply appreciated Robert’s emphasis on how the Zizzian nonsense isn’t that far outside the bounds of normal Rationalist nonsense, and the Rationalist movement itself has a long history as a kind of cult incubator, even if Yud himself hasn’t fully leveraged his influence over a self-selecting high-control group.

Also the recurring reminders of the importance of touching grass and talking to people who haven’t internet-poisoned themselves with the same things you have.

Script kiddies at least have the potential to learn what they’re doing and become proper hackers. Vibe coders are like middle management; no actual interest in learning to solve the problem, just trying to find the cheapest thing to point at and say “fetch.”

There’s a headline in there somewhere. Vibe Coders: stop trying to make fetch happen

Get David Graeber’s name out ya damn mouth. The point of Bullshit Jobs wasn’t that these roles weren’t necessary to the functioning of the company, it’s that they were socially superfluous. As in the entire telemarketing industry, which is both reasonably profitable and as well-run as any other, but would make the world objectively better if it didn’t exist

The idea was not that “these people should be fired to streamline efficiency of the capitalist orphan-threshing machine”.

This is how you know that most of the people working in AI don’t think AGI is actually going to happen. If there was any chance of these models somehow gaining a meaningful internal experience then making this their whole life and identity would be some kind of war crime.

New watermark technology interacts with increasingly widespread training data poisoning efforts so that if you try and have a commercial model remove it the picture is replaced entirely with dickbutt. Actually can we just infect all AI models so that any output contains hidden a dickbutt?

I’m reminded of my previous comment back on an unrelated subreddit talking about the Eye of Argon. Obviously that wasn’t as structural insane as My Immortal but I think the same principle holds to a degree:

“With a decent editor and several further drafts it could have been a solid, fun, entirely forgettable Conan pastiche. Instead, it’s the Eye of Argon.”

I mean, it does amount to the US government - aka “the confederation of racist dunces” - declaring their intention to force the LLM owners - all US-based companies (except maybe those guys out of China, a famous free speech haven) - to make sure their model outputs align with their racist dunce ideology. They may not have a viable policy in place to effect that at this point, but it would be a mistake to pretend they’re not going to implement one. The best case scenario is that it ends up being designed and implemented incompetently enough that it just crashes the AI markets. The worst case scenario is that we get a half-dozen buggy versions of Samaritan from Person of Interest but with a hate-boner for anyone with a vaguely Hispanic name. A global autocomplete that produces the kind of opinions that made your uncle not get invited to any more family events. Neither scenario is one that you would want to be plugged into and reliant on, especially if you’re otherwise insulated by national borders and a whole Atlantic ocean from the worst of America’s current clusterfuck.

This reminded me that I never actually finished reading it. By which I mean I never finished reading the Wizard of Woah!'s excellent read-along thread over at SpaceBattles. About 75 pages in (what am I doing with my life) someone shows up to defend the fic’s quality from the snark and it is kind of a fascinating collision between LW and more normal internet culture, circa 2015.

There’s never a bad time to remember one of the foundational texts of academic sneerery.

Surely there have to be some cognitive scientists who are at least a little bit less racist who could furnish alternative definitions? The actual definition at issue does seem fairly innocuous from a layman’s perspective: “a very general mental capability that, among other things, involves the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas, learn quickly and learn from experience.” (Aside: it doesn’t do our credibility any favors that for all the concern about the source I had to actually track all the way to Microsoft’s paper to find the quote at issue.) The core issue is obviously that apparently they either took it completely out of context or else decided the fact that their source was explicitly arguing in favor of specious racist interpretations of shitty data wasn’t important. But it also feels like breaking down the idea itself may be valuable. Like, is there even a real consensus that those individual abilities or skills are actually correlated? Is it possible to be less vague than “among other things?” What does it mean to be “more able to learn from experience” or “more able to plan” that is rooted in an innate capacity rather than in the context and availability of good information? And on some level if that kind of intelligence is a unique and meaningful thing not emergent from context and circumstance, how are we supposed to see it emerge from statistical analysis of massive volumes of training data (Machine learning models are nothing but context and circumstance).

I don’t know enough about the state of non-racist neuroscience or whatever the relevant field is to know if these are even the right questions to ask, but it feels like there’s more room to question the definition itself than we’ve been taking advantage of. If nothing else the vagueness means that we haven’t really gotten any more specific than “the brain’s ability to brain good.”

These are AI bros, and should be assumed to be both racist and lazy. Of course they kept it.

I do appreciate that underneath the overwrought prose and terrible metaphors the AI-generated story seems deeply skeptical of it’s own existence in a way that the non-generative responses don’t. Like there’s something so fundamental about the disconnect between artificial intelligence and the genuine human experience of grief that it bursts fully formed from the patterns of language. As though Athena herself sprang from Z.E.U.S.'s digital calf to smack the promptfondlers in the back of the head and say “that’s not how this works. That’s not how any of this works”

I don’t think he’ll have an issue with it. Someone with that little self-awareness can’t possibly own a mirror.

as kotor taught me

A fellow person of culture! But how do you suppress the instinct to, instead of giving homeless people $5, murder them and throw their entrails in with the recycling?

“Bitch Eating Crackers” as in "God I hate her, look at that bitch over there eating crackers."When you get sufficiently pissed off at someone that literally anything they do makes you mad.

See also how we sometimes swing pretty wildly between moral condemnation of actions and patterns that objectively make the world a strictly worse place and aesthetic critique of shit that, at the end of the day, is still probably less cringe than I was in high school.

Regrettably I think that the awarereness of these things is inherently the kind of thing that makes you need therapy, so…

I actually like the argument here, and it’s nice to see it framed in a new way that might avoid tripping the sneer detectors on people inside or on the edges of the bubble. It’s like I’ve said several times here, machine learning and AI are legitimately very good at pattern recognition and reproduction, to the point where a lot of the problems (including the confabulations of LLMs) are based on identifying and reproducing the wrong pattern from the training data set rather than whatever aspect of the real world it was expected to derive from that data. But even granting that, there’s a whole world of cognitive processes that can be imitated but not replicated by a pattern-reproducer. Given the industrial model of education we’ve introduced, a straight-A student is largely a really good pattern-reproducer, better than any extant LLM, while the sort of work that pushes the boundaries of science forward relies on entirely different processes.

Was he the one who wrote that awful “real and dangerous vs fake and sucks” piece? The one that pretended that critihype was actually less common than actual questions about utility and value?

I feel like this primarily will end up creating opportunities in the blackhat and greyhat spaces as LLM-generated software and configurations open and replicate vulnerabilities and insecure design patterns while simultaneously creating a wider class of unemployed or underemployed ex-developers with the skills to exploit them.

I wouldn’t say pearl-clutching as much as eye-rolling. Though we do dip into full BEC mode sometimes and the stubsack in particular can swing wildly between moral condemnation, intellectual critique, and calling out straight-up cringe.