California won't require big tech firms to test safety of AI after Newsom kills bill
Source: The Guardian
California governor Gavin Newsom on Sunday vetoed a hotly contested artificial intelligence safety bill after the tech industry raised objections. Newsom said that requiring companies to stress test large AI models before releasing them could drive AI businesses from the state and hinder innovation.
California is home to 32 of the worlds 50 leading AI companies, the governor said in a statement accompanying the veto. The bill applies stringent standards to even the most basic functions so long as a large system deploys it. I do not believe this is the best approach to protecting the public from real threats posed by the technology.
The Safe and Secure Innovation for Frontier Artificial Intelligence Models Act, officially known as SB 1047, would have targeted companies developing generative AI which can respond to prompts with fully formed text, images or audio, as well as run repetitive tasks with minimal intervention. Companies building models costing more than $100m would have been required to implement kill switches for their AI as well as publish plans for the testing and mitigation of extreme risks.
-snip-
Blake Montgomery, Johana Bhuiyan and agencies
Sun 29 Sep 2024 23.27 BST
First published on Sun 29 Sep 2024 22.03 BST
Read more: https://www.theguardian.com/us-news/2024/sep/29/california-governor-gavin-newsom-vetoes-ai-safety-bill
PoindexterOglethorpe
(26,727 posts)if something goes horribly wrong, oh well, at least no time was wasted on testing, right?
Dear lord.
SamKnause
(13,802 posts)ancianita
(38,514 posts)On second thought, though, the decision is more about CA minding businesses, and letting the federal government legislate AI because of its national and global impact. Which is actually a better, at scale way to handle AI.
While competition is good in some economic arenas, AI is an extraction industry. It's much like mining and fossil fuel industries,, and as such, there doesn't need to be anti-monopolistic competition around such a global power; there needs to be tight regulation of that power.
AI cannot be allowed to take over as the "brain" of the global Internet nervous system. And it's only big scale federal government power like the US that can make sure this is still a country that is run of, by and for The Humans. The EU, which has been way ahead of the US in regulating global platforms (ask Elon) in the public interest, would agree.
50 Shades Of Blue
(10,887 posts)Response to Eugene (Original post)
Post removed
CentralMass
(15,537 posts)Jk23
(401 posts)This would drive companies and innovation out of California. The cost to follow these rules would allow only the very wealthy to develop AI wiping out small and mid-sized innovation.
It is regulation to shrink the playing field and create monopolies.
I need to come into the now. When I see "AI" my eyes glaze over and I don't have the intestinal fortitude to learn the details. Putting on the to-do list. I need the "AI for Dummies" book.
ananda
(30,812 posts)He keeps vetoing good bills.
Has he turned to the dark side?
Lulu KC
(4,182 posts)He doesn't seem bogus or naive enough to change direction in the interest of the presidency. He is who he is and does he actually think he can change that image? I need a good longitudinal analysis of him to quench my curiosity.
ananda
(30,812 posts)All the AI stuff, everywhere, really bothers me.
I feel for all the actors and artists put out of work.
bahboo
(16,953 posts)it's usually pretty solid...and progressive in terms of achieving goals in the long run...a pragmatist he is...
OldBaldy1701E
(6,337 posts)(Sci-fi has this weird ability to predict things when it comes to technology. It is more accurate that we sometimes want to believe. You can scoff if you want. I am sure that people scoffed when they were told that machines could do factory work as well if not better than humans can. Let's go ask those laid off folks what they think about how ridiculous that idea sounded at the time.)
Blue Full Moon
(1,153 posts)From Stephen Hawking, "The development of full artificial intelligence could spell the end of the human race." "It would take off on it's own, and re-design itself at an ever increasing rate." "Humans, who are limited by slow biological evolution, couldn't complete, and would be superseded."
To quote from a video game, " the cake is a lie."
Novacene: The Coming Age of Hyperintelligence by James Lovelock is an insightful read.
Now there is research pointing out that AI tweaks to photos and videos can alter memories. It causes people to misremember what they have seen.
So yes it does need strict regulations. And Newsome is trying to curry favor.
TexasBushwhacker
(20,667 posts)rather than drive it away to states that are far less regulated.
Blue Full Moon
(1,153 posts)This is to gain support from tech bros because I think, he wants to be president and he needs their financial support. He has been vetoing his own parties bills. He wanted to run for president and that was why he was one of the loudest voices to remove Biden.
But my point is that is has to have strict controls. Seems like this issue is flying under radar and probably the biggest threat of all.
TexasBushwhacker
(20,667 posts)JMHO, state laws and regulations have their place, but something as far reaching as AI probably needs regulation on a national level.