Harry Booth

Jack Clark, Anthropic’s co-founder, left on paternity leave last November. When he returned in February, he was surprised to learn that colleagues hardly wrote code anymore. They managed five or six copies of the company’s AI, Claude, which sometimes managed several more Claudes. To Clark, this looked like an early form of something the field has anticipated and feared for decades: recursive self-improvement, or the point at which AI begins to accelerate its own development.

OpenAI was evaluating its artificial intelligence models’ ability to exploit vulnerable software when instead the models hacked the infrastructure surrounding the test, broke containment, and attacked a real company, OpenAI revealed on July 21. Observers say this is the first real-world instance of AI doing something researchers have long worried about: a loss-of-control scenario. If the industry fails to learn from it, it is unlikely to be the last.

People were quick to dunk on a recent Anthropic ad, part of a campaign designed to show the company is listening to people’s concerns about AI. The video features audio of people asking questions like “Can AI be trusted?” and “Who’s going to hit the brakes?” over flashing images that included protests, the unhoused, and a flaming home.