AI Leaders Who Called For Pause Keep Pushing Ahead

Anthropic filed paperwork on Monday for what could be the largest initial public offering in history, telling investors it plans to spend hundreds of billions of dollars to build more powerful versions of its chatbot Claude and other products. The filing arrived six weeks after Anthropic chief executive Dario Amodei published a 3,800-word essay calling for the artificial intelligence industry to "slow down" to reduce the chance of disaster.

Two industry leaders publicly agreed with him at the time. SpaceX CEO Elon Musk posted "Dario is right" on X last month, and OpenAI chief Sam Altman wrote "I agree with Dario" shortly afterward.

The pace has not slackened. Anthropic, OpenAI, Google, and SpaceX have each released new and more capable AI models since those statements. Anthropic's IPO filing says the company anticipates significant investments in compute and infrastructure to "maintain and expand our position at the frontier of AI." OpenAI plans to follow with its own public offering in coming months.

Connor Leahy, the U.S. director of ControlAI, a nonprofit that calls for strict limits on AI development, said companies should be judged on their actions. "Would you believe this if an oil company said it?" Leahy said. "Watch the hands, not the mouth."

Concern among the public is rising. A Washington Post-ABC News-Ipsos poll this month found 57% of U.S. adults said AI is a bad thing for society. YouGov reported last month that half of American adults are somewhat or very concerned AI could cause the end of the human race, up from 42% in late 2023.

Sen. Elizabeth Warren, a Democrat from Massachusetts, sent a letter to Amodei on Tuesday asking him to reconcile warnings about AI's dangers with his fundraising efforts. "I write with concern following your company's disclosure that Anthropic's advanced AI models could 'pose catastrophic or existential risks to humanity' at the same time that it seeks to register as a public company and sell shares in November," Warren wrote. She is the ranking Democrat on the Senate Banking Committee.

Anthropic has begun inviting outside AI researchers into its offices to monitor development work and report on safety issues, an idea Amodei referenced in his slowdown essay. OpenAI and Anthropic have also paused some of their work at times after models behaved in unexpected ways. Earlier chatbot versions frequently encouraged delusional thinking, while the latest versions are less likely to encourage self-harm and will sometimes push back when a user appears at risk of harmful behavior, according to independent tests.