Former Anthropic AI researcher Jacob Coxon on Sunday called for Congress to allow the nation’s leading frontier AI labs to regulate themselves while lawmakers come up with a long-term regulatory framework for slowing the pace of AI development.“My personal opinion would be to insist and allow that the current labs regulate themselves.I do think that the people running the labs, especially their recent communications, are completely genuine.
They would like to slow themselves down,” Coxon told NBC News’ “Meet the Press.”He added that current legislation that’s been introduced in Congress, including the framework for a national “kill switch” if AI starts to exceed the bounds of human control, will quickly be outpaced.A “kill switch probably would work on a lot of AIs for now,” Coxon said.“Although I think it’s important to remember, it’d be quite a lot of switches.
We have a lot of — our data centers are pretty big.It’s not the easiest thing in the world, but definitely doable while it’s constrained to one sort of physical location.”He added that one scenario where a kill switch wouldn’t work was described this week by Anthropic AI CEO Dario Amodei.“Like Dario said, very soon there’s the possibility that the kill switch just wouldn’t work because a swarm might have gone on an internetwide hacking run,” Coxon said.He added that any national laws regulating AI would have to be nimble and able to adapt to new models.“It’s a continuous effort to ensure that every new mind that we create is appropriately, kind of, constrained,” the former AI researcher said.Coxon last week set off a wave of concern about the speed of AI development at the country’s leading AI labs including Anthropic, OpenAI and Google DeepMind.In an X post Tuesday that he published hours after resigning from his role as a researcher at Anthropic, he wrote: “I spent the last three years doing pretraining research at both OpenAI and Anthropic.
Neith...