Encord Explores Brain Waves to Enhance Physical AI Training
Encord, a company specializing in data tools for AI models, is experimenting with using brain wave data to improve the training of physical AI systems. In a warehouse in San Leandro, California, Encord is collaborating with Zander Labs to collect brain wave data from operators performing tasks like playing Jenga. This data is intended to enhance the training of robotic models by providing insights into mental states such as error and intent. The initiative is part of a broader effort to address the scarcity of real-world physical training data, which is seen as a bottleneck in the development of advanced robotics.