I write about AI for a living. Lately, I don’t know what to tell people
At the same time, we’re now being told that AI may, quite literally, kill us all within the next decade, with the titans of AI furiously penning joint letters about the dangers of rogue and self-improving AI models while calling for a global slowdown in AI development.
OpenAI’s Sam Altman, for one, writes of the need for “a federal framework that sets consistent safety requirements for frontier AI,” while also noting that the “shift to focusing on safe development and evaluation will need new tools,” which sounds suspiciously like more AI.
Meanwhile, Anthropic boss Dario Amodei declared on Sunday the need to “slow the pace at which we improve the capabilities of AI models” amid “dramatically faster” over the past several months, not to mention the scary reports of AI agents going rogue.
But neither Altman nor Amodei think stopping AI development is the answer; rather, “pacing” is the operative word, with Amodei raising the spectre of a “Chinese lead in AI” that “would pose grave danger for the United States and the world.”
China, for its part, has fired back at the AI “pacing” proposals as “fear mongering,” while U.S congressional leaders seem wary of putting the issue on the political table.
In the middle of all this is the rest of us, who’ve had AI suddenly foisted on our lives and told to use it, or face the consequences.
