Machenta unveils robot model claimed to learn tasks from one demonstration

Machenta says its new robot model can learn a task from one demonstration, without further training, and was trained entirely in simulation. It announced a $40 million seed round co-led by Creandum and USV. In a later post, the announcer said the model also works on SO-101 robots. A separate observer called instant, in-context learning exciting but said it still needs considerable work.