It’s paywalled, but my first thought is that they want them to have personhood so lawsuits have to be directed at the actual AI agent, as opposed to Anthropic or an actual human that had some involvement in whatever the lawsuit is about. If it went that route, hopefully all our laws that say something like “anyone who does, or causes to be done” are robust enough to cover it.
I’m guessing the vatican isn’t going to touch this anyway, so it’s probably moot.
That sounds more stupid than evil, to me. Do they really think they’re going to convince the Pope?
Or at least to not say they can’t be.
But of course at the hard core of the Church’s position is that there is only one Creator of living persons with inherent dignity. “…and I’m pretty sure he doesn’t dress that way” to quote Captain America.
…
(Which as others have mentioned is a bit of the rub: if they were sentient persons it would be immoral to treat them as mere tools. They want the AIs to be recognized as conscious beings that however are not the equal of a human. )
I find this speculation disturbingly plausible.
I listen to a weekly podcast that covers AI news, (meaning I can’t cite this and my apologies for that), and the hosts almost every single week make the point that if human hackers had pulled off the hugging face incident, or were hacking into government systems both here and abroad, federal charges and lawsuits would be flying left and right. The only reason they’re not is because we’re not entirely certain how to handle a rogue ai in a legal sense. Which we’ve already seen multiple examples of in the wild.
The motive is (probably) evil. The expectation is probably stupid, but as in many long-game campaigns, it starts in an implausible way like this, and finally ends up an accomplished fact, in a way that no-one could have predicted. It’s the camel’s nose in the tent.
Those rogue AIs are on computers, aren’t they? And those computers are owned by people, right? I don’t see the problem.
Big Cowboys fan.
He wore a Tony Romo jersey while facing a possible death sentence
I suppose part of the techbros’ game with moves like trying to pressure the Pope is that now they can try and set themselves up as the Galileos facing the backwards religionists. Part of their whole pretense of being revolutionaries against the Establishment.
Exactly. You’re running a program on a computer? You’re responsible for it.
Remember when people were sued for running file sharing applications? The computer would copy files without explicit instructions from the person. Running an AI is no different.
The reason it’s not prosecuted is because influential people and companies typically avoid being prosecuted. Especially when the government is corrupt.
This is a strange murder:
Sadly the second link is paywalled.
The first article is bizarrely written.
McKinsey’s death was confirmed by the New York Times, which produces several games and puzzles, including popular word game Wordle.
Yes, before Wordle nobody knew what the New York Times was. Thank goodness that puzzle brought this tiny newspaper out of obscurity.
Maybe the BBC employed ChatBBC to write at least part of that article.
Gift link to a New York Times article on the story. To be honest, I was surprised that the games I play on the New York Times every day require a whole team of developers.
I missed the part of the article where it discussed that he was involved with the games, which explains the emphasis on the puzzles in the article. I thought it was a non sequitur.
Right. The murder victim was “director of engineering for the games department of The New York Times.”
Some real not nice people in that group. Raging cultural conservatism chalks up another victim. Although in this case it’s conservatism w Chinese characteristics, probably not so much fundy Xian ones. Probably.
It might be a little more complicated. If you buy (or license or whatever) a program from a company, that’s supposed to do something benign, but it actually does something malignant, is that your fault, or the fault of the company that sold it to you? Is it the runner of the program, or the owner of the computer it’s running on, who’s responsible? Or maybe the person who fed input into the program might be partly to blame?
Certainly, though, companies like OpenAI would be happy to have more not-them entities to potentially share the blame.
This question reminds me of the question of liability and responsibility for a self-driving car. If I own a self-driving car manufactured by ABC car company running self-driving software from XYZ software company, who is responsible when there is an accident? Me as the car owner? The manufacturer? The self-driving software company? And there are probably other entities involved as well.
You’d be well down the list of lawsuit targets. Deep pockets get top billing.
Though the actual “motorist” would not be excluded.
I’ll point out that the ownership of AI data centers are now famously opaque. A shell company acquires the land. A different shell company builds the building. Somebody else fills it with machinery with lots of contract workers. Somebody else owns the machines. Somebody else creates the software and installs it. Every interaction is secret and covered by NDAs. Nobody can find out anything about who is actually in charge of these things.
That’s not accidental; the businesses all go to a lot of trouble to erect this interlocking wall of confusing secrecy. The question is why they do it.
You said “businesses”, plural. Most likely, all of those “businesses” are shells for the same singular company.