Anthropic Report Reveals AI is Accelerating Its Own Development Speed
A new report from the AI safety organization METR reveals that Anthropic is using its own AI models to significantly accelerate research and development. According to the 11-week investigation, Anthropic's model improvements are occurring roughly twice as fast as they were a year ago, effectively compressing two years of historical progress into a single year. The report highlights that this acceleration is driven primarily by AI assistance rather than just increased computing power. Internal employees reported productivity gains of up to 12 times when using the company's most capable internal models. While current gains stem from human-AI collaboration, researchers warned that signs of fully autonomous AI contributions are emerging. These findings arrive as Anthropic prepares for a potential public offering and amid ongoing debates regarding AI safety. The report also noted that Anthropic utilized redaction rights extensively, leaving some questions about the full extent of risks and the capabilities of unreleased internal models.
Topics: AI development acceleration AI safety research Anthropic automation METR report
Read the full article at The Verge
Part of the story Anthropic Report Reveals AI is Accelerating Its Own Development Speed (2 articles)