Google Develops Smarter Robots Using Advanced AI Language Models
- Google created a new robotics model called RT-2 that integrates AI language technology to improve reasoning and problem-solving.
- RT-2 allows robots to understand tasks like throwing out trash by training on images and text from the internet.
- In tests, RT-2 nearly doubled robots' performance on new scenarios compared to previous models.
- Robots using RT-2 can follow commands and make basic inferences to choose appropriate actions.
- While not yet commercially available, this technology could have future applications in warehouses, healthcare, and homes.