Conversation with Ishan Mizra on Self-Supervised Learning and AI
·2h 30m
In this engaging discussion, Ishan Mizra, a research scientist at Facebook AI Research, shares insights on self-supervised learning in the context of machine learning, particularly in computer vision. He explains different types of learning paradigms, their significance, and the innovative methods of self-supervised learning that draw from large datasets. The conversation covers the challenges and potentials of multimodal learning, the prospects of autonomous driving, and the intersections between AI, human interaction, and broader philosophical questions about meaning and consciousness
Topics
AI
machine learning
self-supervised learning
computer vision
multimodal learning
autonomous driving
deep learning