Robots doing unexpectedly clever things: problems unsolvable six months ago now cracked
tomssilver · x · 2026-09-25
Tomás Silva (with Nishanth Kumar adding context) shared an observation from a robotics project: he's seen robots do plenty of unexpected things, which usually means it's time to fix a bug — but this new project is one of only a few times in his career where robots did things that were unexpectedly clever, in ways nobody on the team anticipated.
Kumar added that a number of these problems were unsolvable, or at least very hard to write down a solution for, roughly six months ago — a striking pace of progress. First-hand notes like this from robotics researchers are often read as signals of jumps in embodied generalization.
More from Embodied
- Tesla ramps Optimus robot production ~10x, Musk reportedly eyes Grok as the brain — McDonaghMatthew · 2026-09-26
- Two years on, π0.5 remains the only consensus open-source VLA, and that's a problem — paigeinsf · 2026-09-26
- Perceptron releases Mk1.5, one model to fly drones and control robots — AkshatS07 · 2026-09-25
- ROS 2 Makoa Charts a Less-is-More Path: Cutting, Not Adding — facontidavide · 2026-09-25
- Sims and Sensors Called the Most Fertile Ground at the Infra Layer — pzakin · 2026-09-25
- Robot simulation is the cleanest anti-gaming incentive, argues Bittensor SN49 — bittingthembits · 2026-09-25