In this episode of the Big Technology Podcast, Alex Kantrowitz interviews Demis Hassabis, CEO of DeepMind, and Sergey Brin, co-founder of Google, about the frontiers of AI. The discussion covers the potential for improvement in frontier models, the importance of both scale and algorithmic advances, and the need for more data centers. They delve into the concept of "thinking" in AI, as demonstrated by DeepMind's DeepThink, and its potential to bring the industry closer to AGI. Hassabis defines AGI as a system capable of the range of intellectual feats of the best humans, while Brin anticipates AGI being reached before 2030. The conversation touches on the role of emotion in AGI, the development of self-improving systems like Alpha Evolve, and Brin's return to Google to focus on AI. They also discuss the potential of AI-powered smart glasses and the challenges of video generation, including model collapse.
