a16z Podcast The Frontier of Spatial Intelligence with Fei-Fei Li
81 snips
Sep 19, 2024 
Fei-Fei Li, a Stanford professor and AI pioneer behind ImageNet, joins Justin Johnson, co-founder of World Labs and expert in generative models. They discuss the evolution from AI winters to today's groundbreaking advancements in deep learning and spatial intelligence. The duo explores how innovations in visual understanding are transforming industries, the interplay of compute power and data, and the exciting future of 3D environments in gaming and education. Their insights reveal the importance of collaboration in pushing the boundaries of AI.
AI Snips
Chapters
Transcript
Episode notes
Key Epochs in AI
- Key epochs in AI often revolve around specific datasets or algorithmic breakthroughs.
- ImageNet enabled computer vision, while transformers and stable diffusion propelled generative AI.
Compute and Data in AI
- Compute is often underestimated in AI's progress, as its growth has been astounding.
- New data sources, often implicitly labeled by humans, also unlock deep learning.
Justin Johnson's PhD Journey
- Justin Johnson's PhD work reflects the field's trajectory, starting with data and moving towards generative models.
- His work on real-time style transfer and early generative image models was pivotal.