Teaching human values to artificial intelligences

Two Cornell experts in artificial intelligence (AI) have joined a nationwide team setting out to ensure that when computers are running the world, they will make decisions compatible with human values. "We are in a period in history when we start using these machines to make judgments," said Bart Selman, professor of computer science. "If decisions are properly structured, the horrors we've seen in the movies won't happen." Selman and Joseph Halpern, professor of computer science, have become co-principal investigators for the Center for Human-Compatible Artificial Intelligence, a nationwide research effort based at the University of California, Berkeley. Initially they will collaborate with scientists at Berkeley and the University of Michigan.