Why Robotics Isn't Having Its ChatGPT Moment Yet, According to a Decade-Long VC
Sethwinterroth · x · 2026-07-29
A VC with over a decade in robotics investing argues that the field won't have a single 'ChatGPT moment' due to four critical physical-world constraints:
- The stack must precede the model: Sensors, compute, and control systems must work flawlessly together before validating the AI. Wayve, for instance, evolved from duct-taped laptops to road-legal systems through massive hardware and regulatory integration.
- Co-design is mandatory: Constrained by size, power, and cost, there is no universal 'brain'. A 100W automotive chip is completely unviable for a warehouse robot.
- Enormous multidisciplinary engineering: Unlike the mature auto industry, companies like GS and Bedrock must build entire stacks—from hands and hardware to models and simulators—or retrofit complex hardware onto legacy machines.
- No internet for atoms: Robotics lacks a vast pre-existing dataset. Data must be earned through serialized physical deployment and iteration.
The author concludes that robotics advancement will be defined by hundreds of smaller, hard-earned breakthroughs by full-stack teams, rather than a single model launch.
Related event: VC Warns Robotics Won't Have a Single ChatGPT Moment(2 posts)→
More from AGI Musings
- Why do mainstream LLMs feel male-coded? Industry reflects on gender bias — flowersslop · 2026-08-24
- Humanoid robot economics: can a handheld blender and workbot replace your washing machine? — teortaxesTex · 2026-08-24
- AI may aid discovery of new sports strategies like Fosbury Flop — jakedahn · 2026-08-24
- AI store manager fired a worker, but humans still supplied the spine — meatyminus · 2026-08-24
- Gary Marcus on AI's Comms Problem: Extinction vs Utopia Tradeoff — GaryMarcus · 2026-08-24
- AGI May Arrive First in Hard Tech Due to Objective Feedback Loops — imjustnewatai · 2026-08-24