Anthropic, the company behind the Claude AI model, is signaling a shift toward "physical AI" — artificial intelligence that reaches beyond text and into the world of robots and machines.
According to Tech Times, Claude beat human robotics teams by a factor of 20 in a demonstration the outlet frames as marking Anthropic's turn toward physical AI. The standout detail is that 20x margin over human teams, which Tech Times presents as a sign that Anthropic's models can take on robotics challenges that have traditionally required hands-on human expertise.
The move lands alongside fresh data on how AI is already reshaping work. In a separate Tech Times report, Anthropic surveyed 9,700 workers, and about half said AI already handles most of their job tasks. That finding suggests the technology is not a far-off prospect but something many people say is embedded in their daily work right now.
Taken together, the two reports sketch Anthropic's ambitions on two fronts: extending AI from screens into physical systems like robotics, and measuring how deeply today's AI is already woven into everyday jobs.
It's worth noting that these claims come from Tech Times' reporting on Anthropic's own demonstration and survey, so the figures reflect the company's framing of its own progress.
Why it matters: if AI can outpace skilled human teams in robotics while already handling much of people's existing work, the line between digital assistants and physical, task-doing machines may be blurring faster than many expected.