The measure, aimed at reducing potential risks created by AI, would have required companies to test their models and publicly disclose their safety protocols to prevent the models from being manipulated to, for example, wipe out the state’s electric grid or help build chemical weapons.
How exactly do LLMs do that? If you've given an LLM's pseudorandom output control over your electrical grid, no regulation will mitigate your stupidity.
Could he understand the halting problem? I doubt he does, but the legislators evidently don't either
I think it's more about asking it the steps to create a bomb or how to disrupt the grid, for example, where to cut the major edges.
asking it the steps to create a bomb
That sounds like a self-correcting issue right there
That, and the Internet has been teaching people how to create bombs since the dial-up days. I don't predict that LLM's will be either a benefit or a detriment to that particular strain of natural selection.
anyone remember the anarchist's cookbook?
Still a public safety issue.
Is it more of a public safety issue than if they actually build a working one from a legit bomb manual and deploy it?
No, but I think it could make the knowledge more easily available which increases the risk that it may happen.
I see you've never heard of the Anarchist's Cookbook
I think I heard about it before, but instead of having to remember that, I could just ask an uncensored LLM.
The actual point was, bomb making instructions have been floating around on search engine results since the days of dial up. That particular manuscript itself has existed since before the days of the Internet. There's nothing cgpt could give you that you couldn't have found by typing the same query into Google. Getting the instructions is literally the easiest, least effort, least risk part of building a bomb.
How exactly do LLMs do that?
If you hook an LLM up as an interface replacement for a manual/analog Power Plant interface and start asking the translator to intuit decisions based on fuzzy inputs, you can create a cascade of errors that result in grid failure.
If you’ve given an LLM’s pseudorandom output control over your electrical grid, no regulation will mitigate your stupidity.
This rule would prevent a business or public regulator from doing such a thing without proving out safeguards.
And the governor vetoed it.
Good.
All this bill would have done is given OpenAI/Anthropic and such an effective monopoly (and probably destroy the planet with their insane scaling schemes) by destroying the open model ecosystem. I think fediverse vs. corporate social media is a good analogy, and this is kinda like sniping the Fediverse because it's "too dangerous" if it gets too big, without actually being specific on how to deal with that, but actually sniping it because its a competitive threat.
And yes, OpenAI opposed this, but that was lip service. Don't believe a word that comes out of Altman's mouth.
Newsom on Sunday instead announced that the state will partner with several industry experts, including AI pioneer Fei-Fei Li, to develop guardrails around powerful AI models.
That’s reassuring—Li is one of the best-qualified people for the role, and she isn’t in the pocket of any of the major players.
Nancy Pelosi, argued that the bill would “kill California tech” and stifle innovation.
As long as the critics of a safety regulation need nothing better than such stupid, short-sighted arguments, nobody will ever be safe.
Trying to do the same thing in EU I guess. It's funny how the tech giants are mad at it and not releasing their latest energy black hole data pumps in EU. It's like cocaine gangs threatening us to not sell in our countries if we don't change the laws. No, thanks.
Meta: I’ve noticed a lot of VOA links on Lemmy lately, and I’d like to understand why. As I understand it, VOA is essentially a national propaganda news organization targeting an international audience (similar to RT). Why is that a good source for article sharing? Especially in the case of the article at hand, which is just a VOA republication of an Associated Press piece that could have been linked originally.
Fun fact, Gavin Nelson is actually a living AI. He doesn't always get the answer right but he does always have words that sound like a plausible answer. He also creates artwork in his mind.
Gavin Nelson, that's the guy that did "After the Rain," right? Love that guy. Fitting song for these trying times.
Ah yes, the sequel to "Are Those Rain Clouds?" Thought provoking films.
Is that Major Nelson's first name?
Gruesome Newsome's veto pen is to the right of Reagan's.
This is a most excellent place for technology news and articles.
