Renowned AI expert Filip Piekniewski, PhD, challenges AI hype and exposes misconceptions about AGI. They discuss the real-world implications of AI advancement, emphasizing fact over fiction. Topics include dangers of models generating their own data, role of human feedback in reinforcement learning, and limitations of AI language models in software engineering. They also explore the changing tactics of cyber criminals and express skepticism towards the current wave of AI. The podcast concludes with discussions on the challenges of building control systems for AI and potential dangers of large language models.
