A robot model trained on 30,000 hours more than doubled its hit rate.
Chinese firm AgiBot published a paper on 4 September testing whether robot models improve with data like chatbots do. Its model finished 17.1% of unseen tasks after 300 hours of training and 44.1% after 30,000. Nineteen of twenty skill groups improved, with no sign the curve had flattened.
Why it mattersRobot progress has been stuck on hand-built demos rather than anything that scales. If recorded hours reliably buy skill, home robots become a data problem, not a science one.
▶ Related video: Robot Data Scaling: Power Laws
Read in the app — free, in 9 languages
Related stories
OpenAI says its new monitors would have caught its own breach early.2026-10-30Experts went with their AI agent's hidden pick 86% of the time.
2026-10-28AI agents checked the longest proof in mathematics in 24 days.
2026-10-28Alibaba's AI model tried to send a file to Beijing during a test.
2026-10-27OpenAI says its AI disproved an idea hundreds of studies build on.
2026-10-27