"The Cognitive Revolution" | AI Builders, Researchers, and Live Player Analysis Interfacing with AI, with Linus Lee of Notion
67 snips
Apr 15, 2024 Linus Lee, AI product leader at Notion, shares his innovative projects in exploring AI systems and interfaces. He discusses the transformative potential of language models to enhance human abilities instead of replacing them. Linus highlights ideas around advanced interfaces, the Refusion project blending sound and vision, and the interplay between AI and human creativity. He delves into the mechanics of AI cognition, the future of AI reasoning, and the importance of design to maintain human agency through evolving technologies.
AI Snips
Chapters
Transcript
Episode notes
Learning by Autopsy
- Linus learned about deep learning by dissecting open-source models like GPT-2 with print statements.
- He progressed to frontier research by reading papers and writing implementations for niche problems.
How Humans (and AIs) Think
- Human brains might construct narratives to explain actions retroactively, not the other way around.
- Language models work by statistically modelling probability distributions over possible outputs and sampling.
Beyond Reasoning
- Most documents are interpolations of existing ones, so focus on instruction following rather than pure reasoning.
- Improved speed and cost-efficiency are crucial for handling large-scale text generation.