Google and Tesla are conducting similar research, with robots learning from experience to perform tasks. However, the process is often slow and labor-intensive, and providing sufficient training data is challenging. Despite these challenges, TRI researchers have trained robots in over 60 complex skills and aim to increase this to 1,000 by the end of 2024.
Key takeaways:
- Toyota Research Institute (TRI) has used generative AI to teach robots how to perform tasks such as making breakfast, by giving them a sense of touch and showing them how to do tasks.
- The researchers are attempting to create "Large Behavior Models" (LBMs) for robots, which learn by observation and can perform new skills they've never been taught.
- Over 60 challenging skills have been taught to the robots, with the aim to increase this to 1,000 by the end of 2024.
- Google and Tesla are conducting similar research, with the aim for AI-trained robots to eventually carry out tasks with little to no instruction, similar to a human being.