Dario Amodei spent 2,000 words laying out to a critic on X why Anthropic must persist in the AI ‘race to the top’

Anthropic CEO Dario Amodei disputed a criticism that the business was putting AI safety in the rear place in its "race to the top."

On Tuesday, the administrator wrote a nearly 2,000-word reply to a crucial X article which said Anthropic was trying to win a competition to superintelligence and deprioritizing AI safety.

The first Monday X article was by Nathan Calvin, the chief scheme authoritative of Encode AI, an AI-safety assistance nonprofit. Calvin listed nine actions by Anthropic complete the former twelvemonth that he stated undermined its safety commitments, including signing deals for additional than $500 milliard in compute and increasingly relying on Claude to create its AI models. His article was lengthy, too, at complete 1,500 words.

Amodei stated in his reply that during he was a "fan of the investigative work" Calvin had done for AI safety, he disagreed alongside many of Calvin's criticisms.

Amodei stated his stance was that Anthropic can do additional to keep the AI industry harmless if it stays at the frontier.

"Our scheme has continually been what I've called the 'race to the top' — by showing how it's imaginable to build the innovation responsibly, we can lift the bar for the complete industry," Amodei wrote.

He afterward wrote out his own nine-item catalog of actions that were "very costly" to Anthropic commercially, but had "strong safety benefits." This included releasing its mighty Mythos example lone to a choose collection of cybersecurity partners and restricting the transaction of its AI services to Chinese state-controlled companies.

However, Amodei acknowledged Anthropic could have handled several of the issues Calvin raised better. These included a draft invoice Anthropic circulated to Congress this duration — obtained exclusively by Politico — that critics stated would create it harder for the national authorities to obstacle hazardous AI models.

"We'll continually try to do improved but, again, I am extremely arrogant of our track document and remain by it," he wrote.

Amodei and Calvin's toggle comes at a period whenever example safety has rotate into among the most pressing issues in the AI sphere. In September, AI researchers predicted an "AI apocalypse," in which the increasingly capable and rogue innovation could slay all of humanity in the next decade.

In reply to the concerns, AI chiefs akin Amodei, OpenAI CEO Sam Altman, and SpaceXAI CEO Elon Musk all accepted to "pace the frontier," referring to slowing downward AI betterment to justify there were adequate checks and balances in place.

Meanwhile, Anthropic is on the brink of its hotly anticipated IPO, expected to catalog at a $2 trillion evaluation in November. Last week, it reported a third-quarter income of $16.8 billion.