SRI International selected by DARPA to develop AI system that continually learns

SRI International selected by DARPA to develop AI system that continually learns

SHARE

US: SRI International has been selected by the Defense Advanced Research Projects Agency (DARPA) to develop a next-generation artificial intelligence (AI) system able to learn continuously and apply that learning to become better and more reliable at performing new tasks. The contract will be supported under DARPA’s Lifelong Learning Machines (L2M) Program.

 While AI systems have become core to many commercial and government applications, they are not able to handle new scenarios that they are not trained on. AI systems today can repeatedly make the same mistakes. Even with retraining, today’s systems are prone to “catastrophic forgetting” when a new item disrupts previously learned knowledge.

“Our goal is to address these limitations by enabling AI systems to know what to learn and when,” explained Sek Chai, Ph.D., technical director in the Center for Vision Technologies, SRI International. “Since memory is a key element of cognitive function, our research is focused on understanding and applying biological memory transfers to new AI algorithms that can fundamentally improve their performance throughout their fielded lifetime experience.”

Biological memory transfer is a complex sequence of dynamic processes, with local and global synchronization patterns. These processes support memories with flexibility in expression for future thinking, foresight, planning, and creativity. As part of the L2M program, SRI researchers will develop AI algorithms based on biological mechanisms in memory consolidation and replay. With this AI technology, future cognitive systems such as autonomous robots may be able to continually learn after initial deployment, improving execution performance and overall safety.