Anthropic says Claude leads 26% of R&D, up from 1%
Anthropic, an AI company from San Francisco, just shared that its AI system, Claude, now leads 26% of its artificial intelligence research and development work inside the company, way up from only 1% in March.
In August, more than 90% of the research work happening at the company saw humans and AIs teaming up.
Last August Anthropic ran 30,000 agents
Last August alone, about 30,000 AI agents worked on research and engineering tasks at Anthropic.
Every move made by these AIs gets reviewed to catch mistakes early. On average, one out of every 47,000 actions is stopped.
The company is also putting more computer power toward safety: during a sample week in July, about 6% of the computing power Anthropic used for AI research went to safety work, with the figure rising to 12% for research carried out by AI itself, to make sure things stay responsible as AI's role grows.