
Lex Fridman Podcast #162 – Jim Keller: The Future of Computing, AI, Life, and Consciousness
107 snips
Feb 18, 2021 Jim Keller, a legendary microprocessor engineer with stints at AMD, Apple, Tesla, and more, dives into the future of computing and AI. He discusses the significance of empathy in understanding human behavior and how it influences creativity. The conversation reveals insights on the balance of craftsmanship with theoretical innovation in technology. Jim also explores the challenges in neural networks versus human cognition, the impact of love on success, and the potential for ubiquitous computing to reshape our relationship with technology.
AI Snips
Chapters
Transcript
Episode notes
Tenstorrent's Approach
- Tenstorrent aims to build hardware that natively executes graph programs.
- Its two pillars are Andrej Karpathy's "Software 2.0" and Chris Lattner's MLIR.
Tenstorrent's Goals
- Tenstorrent aims to scale from 100 milliwatts to a megawatt, addressing various problem sizes.
- They prioritize basic functionality and ease of use for their customers.
GPUs vs. AI Hardware
- GPUs, designed for pixel-based programs, may not be ideal for graph-based AI computations.
- Hardware optimized for specific operations like rasterization can significantly outperform software.

