Latest Interviews
Showing 1–3 of 3 interview transcripts.
Clear all filters- Lex Fridman2h 47m
Yann Lecun: Meta AI, Open Source, Limits of LLMs, AGI & the Future of AI | Lex Fridman Podcast #416
Yann LeCun argues that centralized proprietary AI threatens democracy by controlling global knowledge, advocating instead for open-source systems that prioritize diverse information access. He proposes replacing autoregressive language models with Joint Embedding Predictive Architectures (JEPA) to enable machines to learn intuitive physics and plan through abstract world models rather than predicting raw tokens. LeCun predicts that human-level AI requires a decade of development to achieve robust physical reasoning, emphasizing that intelligence will evolve gradually through iterative safety refinement rather than through uncontrollable autonomous takeovers.
- Lex Fridman2h 45m
Yann LeCun: Dark Matter of Intelligence and Self-Supervised Learning | Lex Fridman Podcast #258
Yann LeCun proposes that self-supervised learning serves as the foundational mechanism for building world models, arguing that this approach mirrors human biological learning far more effectively than current data-hungry supervised methods. While progress in language has been substantial, the field faces significant technical hurdles in applying similar principles to high-dimensional vision, necessitating new architectures to handle uncertainty and causality. LeCun further contends that this paradigm shift is essential for achieving strong AI capable of solving complex scientific problems, though he warns that such capabilities may eventually require addressing the ethical implications of autonomous drives and the potential for machine suffering.
- Lex Fridman1h 46m
Michael I. Jordan: Machine Learning, Recommender Systems, and Future of AI | Lex Fridman Podcast #74
Michael I. Jordan, Lex Fridman, Andrew Ng, Zoubin Ghahramani, Ben Taskar, Yoshua Bengio, Yann LeCun
Michael I. Jordan reframes the current state of artificial intelligence not as the engineering of human-like cognition, but as a nascent discipline focused on building large-scale decision systems, while explicitly rejecting premature claims of deep neurological understanding or full brain-computer integration. He distinguishes his approach from pure prediction by prioritizing decision-making under uncertainty and advocates for a shift from ad-based surveillance economies to direct producer-consumer markets that utilize game theory to align incentives with societal health. Jordan concludes that advancing this field requires a blend of rigorous mathematical frameworks, such as empirical Bayesian methods, and broad humanistic education to cultivate the empathy and collaboration necessary for solving unsolved challenges like natural language understanding.