The future of artificial intelligence lies in developing agents capable of performing economically valuable tasks across various real-world domains, according to Dawn Song, the new vice-president of AI research at Meta Platforms. Song emphasized that the aim is not to replace humans but to enhance their efficiency in significant areas, thereby increasing economic value. As a computer science professor at the University of California, Berkeley, and a co-director of its Centre for Responsible, Decentralised Intelligence, Song is a notable figure in AI security and co-founded the AI safety start-up Virtue AI. She recently announced her move to Meta’s Superintelligence Labs, where she will focus on AI safety and security. This development follows the introduction of a new benchmark by UC Berkeley’s RDI center, known as Agents’ Last Exam (ALE), designed to evaluate AI agents’ performance in over 1,500 economically valuable tasks across 55 industries.
Are AI agents the next frontier for economic value?
