awesome! these are going straight to the list of things I should be reading

nah all good! our writing community isn’t meant to be containment, it’s just a different category with different expectations (mainly that feedback should be constructive, since MoreWrite is for writers posting their own work)

I’ve never watched the show or owned any of the toys (we have gobots at home and they’re the knockoffs you can’t figure out how to transform, and when you finally do a piece breaks off) but the theme song from the commercial’s been stuck in my head for 23 hours

I have an extremely rough draft up on our writing community MoreWrite: here it is. it very much deserves to be restructured and expanded; I wrote this first draft in a hurry, and since then I’ve both gotten some very solid recommendations for prior work in the area to read and cite from @dgerard@awful.systems and I’ve experienced more instances where technofascist methods were used to take control over open source projects away from their communities.

I really should write more about technofascism while I still have electricity, clean water, a relatively unfractured global information network, and there aren’t too many gunshots outside

we’ve got a few Banks fans on our instance so it’s an ongoing conversation, but the most recent posts are here

I enjoy Banks a lot and firmly believe he’d be here sneering with us if that were physically possible; it’s tragic that the rationalists and a number of AI fans have leaned into subverting the meaning of his work.

she is, regrettably, a fascist who reputedly still hides right-wing conspiracy shit in her projects

I unfortunately know that startled feeling well; I used to be a fan of the work she did with sectorlambda and similar projects

there’s so much quantum woo in that article I want to sneer at, but I don’t know anywhere close to enough about quantum physics to do so without showing my entire ass

explaining these things to normal humans is how you turn them into humans with years of psychic damage, and I support that mission wholeheartedly

an attempt at a summary:

the dilbert guy, who believes he can hypnotize women into having sex with him, now also believes he knows a magic incantation to teach hypnosis to a chatbot, and heavily implies the chatbot was able to hypnotize him in turn (presumably into having sex with it)

FAR AI tries to imply that the Go bots are still not merely superhuman, but far superhuman: “this result demonstrates that even far superhuman AI systems can fail catastrophically in surprising ways.” Uh huh. [FAR AI]

fuck it’s so disappointing that everything in this space has to be communicated through the cracked lens of critihype — that even the utterly normal failings of a misengineered system must be misrepresented, in true techfash style, as further proof that the system is powerful and is only one more breakthrough away from perfection

presented without comment:

(via @Colophonscrawl)

and because it feels like it’s worth screaming this into the void in case there’s any marketing assholes reading: fuck yes I’m here to kill the modern web

(I have the feeling that this comes from the same shithead who pushed to include spicy autocomplete in Firefox.)

it definitely reads like the same shithead, but I’ve had them blocked on mastodon for some time so I can’t say for sure if it was for rampant LLMery or for doing the “without advertising the modern web would die and you don’t want that do you” thing advertisers do constantly

First of all. You could simply prompt the LLM to become conscious, and I bet none of you so-called AI skeptics have noticed that Open-AI has NEVER included text like that in any of their system prompts.

This is a weird kind of assertion. First of all. You could make facts a token value in an LLM if you had some pre-calculated truth value for your data set. That’s not how it works now but it’s a weird assertion to make about an unknown new generation of AI.

fuck almighty where do you people pull this absolute horseshit from

What Strawberry apparently is, is a machine that reasons, which is NOT similar to what Open-AI ever claimed ChatGPT ever was.

well shit, you’ve got a long list of supposed AI researchers to tell that to. here, I’ll make sure you’ve got plenty of time to read their back catalog of utterly fucking stupid claims!

what do you mean you don’t like when your package search command is one of several random, probably-unmaintained ecosystem packages that has to very slowly index everything every time nixpkgs updates because it doesn’t have access to the evaluator’s internals?

Evaluation is 5-20% faster than 2.18, depending on which benchmark is in use, thanks to eldritch horrors.

this is awesome

nix flake lock --update-input nixpkgs is now the much more reasonable nix flake update nixpkgs.

but this is making me go “fuck yeah” on the inside. it seems like a small change, but I can’t emphasize enough how frequently this command gets used (for every flake dependency, not just nixpkgs) for how longwinded and non-memorable the old form of it was. it’s kind of fucking incredible how many UX warts Nix has just from the old evaluator’s devs digging in their heels on shit like this.

same energy as “your request could not be processed due to the following error: Success”

absolutely!

I do have a question from the Mastodon side of the Fediverse. Is there a cleaner way I can share links to Lemmy communities?

I’ll do a little bit of experimentation tonight on our test instance and see if there’s a cleaner way to do link-only posts from mastodon! this’ll kill two birds with one stone for me — I needed to run some federation tests as part of an infrastructure upgrade I’m looking to deploy to our main instance. unfortunately, the federation between Lemmy and Mastodon is very limited and janky in a lot of ways so there might not be a cleaner way to do it, but the thread you created looks good on our end.

also, I can’t speak to how the experience for Blind users would be in any of the Lemmy apps or its web frontend, but if any of them end up being a more convenient way to interact with our posts, we can definitely get you set up with an account on our instance if desired.

If you want my Javascript prompt injection I made, DM me because I don’t wanna give LLM developers easy ways to put up input and output guardrails against my prompt injection.

definitely! I will reach out on Mastodon when I get the chance; I don’t remember if Lemmy even attempts to federate DMs between us and Mastodon, but I don’t trust it to do it well if it does.

Genuine question.

So rude, you didn’t answer my question at all.

yeah find me one single instance of someone doing this “genuine question” shit that doesn’t result in the most bad faith interpretation possible of the answers they get

If I’m missing something obvious I’d love it if you told me.

• most security vulnerabilities look like they cause the targeted program to spew gibberish, until they’re crafted into a more targeted attack
• it’s likely that gibberish is the LLM’s training data, where companies are increasingly being encouraged to store sensitive data
• there’s also a trivial resource exhaustion attack where you have one or more LLMs spew garbage until they’ve either exhausted their paid-for allocation of tokens or cost their hosting organization a relative fuckload of cash
• either you knew all of the above already and just came here to be a shithead, or you’re the type of shithead who doesn’t know fuck about computer security but still likes to argue about it
• fuck off