NSF Pursuing AI-Based Safe Learning-Enabled Systems Research

The National Science Foundation’s effort to push research on an artificial intelligence-driven Safe Learning-Enabled Systems program has gained traction, with an NSF official saying that they have received “what appear to be some really fantastic proposals.” 

Michael Littman, NSF Division of Information and Intelligent Systems director, said the SLES research challenge is to develop a completely safe software. According to Littman, the SLES research’s idea is for an AI-enabled program also extending beyond data maintenance to an algorithm that “will hopefully learn to be better, rather than degrade” with more data introduced,  Federal News Network reported.

The NSF is also looking at other AI-based programs, such as an interactive human-centered computing program and a robust intelligence program exploring core algorithms’ limits and power in making AI and machine learning possible, Littman disclosed.

The NSF announced its SLES research request for proposals in February, with Jan. 16, 2024, as its second and final submission deadline.

The Open Philanthropy Foundation, SLES co-sponsor, has recommended a $5 million grant to support the program.