AI models getting harder to monitor, says U.K. AI Security Institute's Joseph Bloom

[The Information] Joseph Bloom, head of model transparency at the U.K. AI Security Institute, said the ability to monitor AI models to detect what they are trying to do and why is increasingly hard to maintain. "In general, we're seeing monitorability—in various respects or from various angles—decrease," Bloom said on the latest episode of The Information's AI Deep Dive, in one of the first on-the-record interviews given by a member of technical staff at AISI. The institute is one of the world's largest government bodies focused on AI, and its work includes evaluating models from leading AI companies.