Anthropic doubles AI development speed despite CEO's slowdown calls

Anthropic is advancing its artificial intelligence systems at about twice the speed compared to the previous year, according to an analysis by the independent nonprofit Model Evaluation and Threat Research (METR), as reported by whec.com, ocregister.com, and lowellsun.com.

This acceleration is occurring despite CEO Dario Amodei's public calls for the AI industry to slow the pace of improving AI capabilities to allow safety measures to catch up. Amodei made these calls in a September essay amid growing concerns about AI risks.

METR conducted its analysis by reviewing Anthropic's internal data and interviewing staff. The research found that AI itself is helping speed up development, with no significant bottlenecks in computing power or data availability.

Sam McCandlish, a co-founder of Anthropic, responded to the METR report by acknowledging that AI is accelerating their research, though the exact amount is uncertain. He also indicated that the company plans to publish its own analysis of AI progress speed within Anthropic soon.

The collaboration with METR began independently of Anthropic's embedded third-party evaluators, which are part of its commitment to pacing AI development responsibly. McCandlish noted that this work illustrates the value of external evaluators.

Anthropic is preparing for a significant initial public offering of stock later this month, which could be the largest ever, as the company navigates ongoing debates about AI's potential risks and regulatory approaches.