Physical AI in a World of Humans
summary
Builds offline stacks like LLMs in a box and an internet in a box for rural and medical use, prototypes remote delivery bots and offline transcription, and runs community labs that pair artists with engineers. Outcomes: a projection-mapped mural and CES demo robots, a sub-millimeter motion capture lab cutting rates from ~$10,000/day to $1,500/day, and three bootstrap businesses reaching self-sufficiency in seven months. The claim is practical: make deep tech usable for everyday people by shipping working systems and localizing models rather than promising vague future benefits.
AI-written summary · about
chapters (11)
- 0:00Introduction and fears of AI
- 2:20Focus on physical AI
- 3:20Building knowledge-sharing spaces
- 5:06LLMs in a box and rural use cases
- 6:08Investing in future of work
- 8:13Artist-tech collaboration projects
- 9:14Frontier Tower hackathon solutions
- 10:37Immersive Commons deep tech
- 12:43Learning Layer Labs model
- 14:07Results and excitement in labs
- 15:09Q&A and closing
related talks
47mDr. Simon Müller | AI Track | Compass AI & Data Summit 2025
18mPhysical AI — When AI Leaves the Digital World
01mhumanoid robots and the future of physical AI
36mThe democratization of technology a world where anyone can use a robot
18mRobotics: why now?
16mReal-World Robotics as the Next Frontier for AI?
14mTowards Unobtrusive Physical AI: Augmenting Everyday Objects with Intelligence and Robotic Moveme...
01mPhysical AI is everywhere