He Walked Away From a $7M Robot Dog Project: As AI Gets Stronger, the Scarce Skill Is Deciding, Not Prompting
yongqianme · x · 2026-09-07
The author recounts walking away from a $7M quadruped robotics project in spring 2025. The customer wanted robot dogs doing geological surveys across real unmapped terrain—rock slopes, riverbeds, dense forest, wetlands, snow—with minimal human intervention. After talking to leading vendors in China, the US, and Europe, he found that while lab and closed-course performance had clearly improved, stability, endurance, perception, and locomotion still couldn't survive acceptance on unmapped mixed terrain. His broader thesis: as AI gets stronger, the scarce skill is not prompting but deciding—including judging which projects and deployment stories don't hold up.
More from AGI Musings
- Compute scale-ups undercut algorithmic RSI, researcher calls on labs to release FLOP numbers — 1a3orn · 2026-09-07
- RSI debate flaw: Astra and Mythos were both compute scale-ups, undercutting algorithm-driven recursive self-improvement — 1a3orn · 2026-09-07
- Yacine wonders when AI-designed chips will be taped out directly for businesses and consumers — yacineMTB · 2026-09-07
- Yacine predicts PCB design will run on open-weight models on consumer multi-GPU rigs within a year — yacineMTB · 2026-09-07
- AI gaming timeline goes viral: 3D in 2 years, on-demand AAA in 4, iRobot-level robots in 8 — nptacek · 2026-09-07
- a16z GP names his three favorite personal agent products right now — lennysan · 2026-09-07