A weekend essay from one of the artificial intelligence industry's most prominent chief executives has reopened a long-running debate inside the AI world: whether the companies building the most powerful models are moving faster than they, or society, can safely manage.
Anthropic CEO Dario Amodei published a roughly 3,800-word essay over the weekend arguing that AI companies should slow the pace of improving their models in order to address safety concerns, a notable stance from the head of a company that has built its business on rapid model development. OpenAI CEO Sam Altman publicly agreed with the substance of Amodei's argument in a post on social media, and Tesla and X owner Elon Musk also voiced agreement, giving the call a rare degree of cross-company alignment among leaders who normally compete aggressively for AI talent, compute and customers.
The essay landed amid a separate, more pointed controversy. A departing Anthropic safety researcher had told colleagues in a resignation message, which then circulated on social media, that they believed there was more than a 10 percent chance that AI could pose an existential risk to humanity, a warning that was quickly amplified by other researchers at both OpenAI and Anthropic calling for an industry-wide slowdown. A former Anthropic researcher separately told the BBC that staff inside the company were, in their words, genuinely frightened about the technology's trajectory.
The timing carries particular weight for Anthropic, which has been positioned for one of the largest technology initial public offerings on record. CNBC reported that OpenAI has ruled out pursuing an IPO of its own this year, even as Altman, Musk and Amodei all publicly warn that AI is moving too fast. Bloomberg reported that market watchers expect the warnings to weigh on chipmaker and AI supply-chain stocks in the near term, since a more cautious approach to model development could crimp the near-term revenue growth investors have been pricing in, though most expect the underlying demand for computing infrastructure to remain strong.
Watch for whether Anthropic, OpenAI, Meta and other major labs follow through with concrete changes to development timelines, and how regulators and investors respond to a rare moment of public agreement among rival AI chief executives.
What exactly did Dario Amodei argue?
In a weekend essay, the Anthropic CEO argued that AI companies should slow the pace of improving their models in order to better address safety concerns.
Did other AI leaders agree?
Yes, OpenAI's Sam Altman and Elon Musk both publicly agreed with the substance of Amodei's call for a slower pace of development.