AI agents that provide ‘economic value’ are next frontier, says Meta AI research chief
UC Berkeley’s RDI centre earlier this month introduced Agents’ Last Exam, a new benchmark that tests how well AI agents perform
“The goal is not to replace humans,” Song told the South China Morning Post last week on the sidelines of the World Economic Forum in Dalian, also known as Summer Davos, days before joining Meta.
“But we want these AI agents to be more effective in these important real-world domains and help humans do this work better and provide more economic value,” she said.
Song, a professor in computer science at the University of California, Berkeley and a co-director of its Centre for Responsible, Decentralised Intelligence (RDI), is a renowned Chinese-American scientist in AI security and also co-founder of enterprise AI safety start-up Virtue AI.
Song said on X on Friday that she, along with “many members of the Virtue AI team”, was joining Meta’s Superintelligence Labs as vice-president of AI research to help shape Meta’s AI safety and security efforts.
UC Berkeley’s RDI centre earlier this month introduced Agents’ Last Exam (ALE), a new benchmark that tests how well AI agents perform in more than 1,500 real-world, “economically valuable” tasks across 55 industries.