State of AI 2026 with Sebastian Raschka, Nathan Lambert, and Lex Fridman

I recently sat down with Lex Fridman and Nathan Lambert for a comprehensive 4.5 h interview to discuss the current state of progress of AI, and what the progress in 2026 look like. We cover the LLM landscape from multiple angles, starting with the geopolitics of the AI race and comparisons between today’s leading LLMs. Then we discuss how these systems are actually trained, scaled, and post-trained. Of course, we also discuss current open vs closed models, coding assistants, tooling, and emerging research directions, and closes with practical advice and a look at AGI timelines, compute, and the long-term implications for industry and society. Yes, there is a lot of ground we covered!