New AI startup could shrink server-sized models for use on iPhones

A new startup is developing techniques to compress large-scale artificial intelligence models so they can run directly on mobile hardware like iPhones. By reducing the computational requirements of these systems, this approach could allow complex generative AI features to function locally on devices rather than relying on cloud-based servers. This shift could improve user privacy and enable offline access to advanced AI capabilities currently limited to high-powered data centers.