Seldon AI cofounder says robotics will hit jobs slower than model-driven automation
luke_drago_ · x · 2026-08-04
Seldon AI cofounder Finn Metz argues that robotics will not blindside the economy the way AI models can.
- Robot deployment is relatively predictable: we roughly know how many robots are being built and shipped, and the current number of humanoids is only around 30,000, with very few doing useful work.
- That makes it easier for businesses and workers to prepare for robotics adoption than for white-collar automation, where a new model release can suddenly change the economics of an entire job category overnight.
- He points to math as an example: once a model crosses a capability threshold, tasks can quickly become more economically valuable when done by AI than by humans.
More from AGI Musings
- AI Researchers Flood In, Expect Sharp Output Kinks Across Domains — nabeelqu · 2026-08-04
- Mathematician: Math's Value Lies in New Concepts, Not Just Proving Theorems — _rockt · 2026-08-04
- A Scaling-Hypothesis Thread Framing Pretraining as the Path to Intelligence — lambdaviking · 2026-08-04
- Musk says source code should vanish, but the post argues software still needs a readable layer — sheakspeares · 2026-08-04
- Waymo CEO says its first self-driving demo took 18 months, product took 15 years — Y Combinator · 2026-08-04
- Stanford panel tackles AI beyond chatbots, from agents to ownership — PolarBearby · 2026-08-04