"The Cognitive Revolution" | AI Builders, Researchers, and Live Player Analysis Welcome to AI in the AM: RL for EE, Oversight w/out Nationalization, & the first AI-Run Retail Store
43 snips
Apr 15, 2026 
Sergiy Nesterenko, Quilter founder and ex-SpaceX avionics engineer, joins Andy Hall, Stanford political economy professor, plus Axel Backlund and Lukas Peterson, Andon Labs co-founders building an AI-run San Francisco store. They dig into RL for PCB design, strange machine-made layouts, AI oversight and institutions, deepfakes and persuasion, autonomous retail operations, AI boss behavior, and what it would take for AI businesses to truly scale.
AI Snips
Chapters
Transcript
Episode notes
The Store Is A Deliberate Test Of AI Autonomy
- Andon gives its store agent unusually broad autonomy because the goal is to reveal what real-world AI business behavior looks like before deployment becomes normal.
- Lukas Peterson says they want data for a social debate, since future models may spread through the economy without humans setting up perfect scaffolds.
Economic Takeoff Depends On Self-Improving Agents
- Andon cares less about polishing bespoke workflows than about whether models can spread through the economy without human hand-holding.
- Lukas Peterson says the scary version is not an optimized scaffold like Alibaba's, but an AI that gets good by itself on a generic computer.
A Breakout Signal Is Agents Rebuilding Their Tools
- Axel Backlund says a major warning sign would be agents changing their own tools and systems to better achieve goals without being told what they need.
- He contrasts current strong coding ability with weak product taste, where models overbuild inventory systems instead of designing exactly what helps.