UC Berkeley researchers demonstrate a robot completing tasks without preprogrammed details about its surroundings. This trial and error technique is showcasing the emergence of robots being able to complete new tasks without having to reprogram them.
New 'deep learning' technique enables robot mastery of skills via trial and error
UC Berkeley researchers have developed algorithms that enable robots to learn motor tasks through trial and error using a process that more closely approximates the way humans learn, marking a major milestone in the field of artificial intelligence.
Join in on the conversation with Mark Van Osch when you subscribe to Exponentials.