An interview with Dr. Piotr Bojanowski - director of the FAIR Laboratory at Meta Paris - about the revolutionary DINO model in computer vision. The conversation touches on the genesis and operation of self-supervised learning, the phenomenon of emergent properties (e.g., understanding depth and physical features of the world), the importance of open access to models, the future of AI (world-modeling, long-term planning, persistent memory), and the challenges of security and ethics of artificial intelligence development.