Robotics engineer dismantles the 'humanoids fit human spaces' debate: it's about safety
KyleMorgenstein · x · 2026-11-20
- Robotics engineer Kyle Morgenstein responded to cbames' claim that humanoid companies' 'human form fits existing infrastructure' pitch is bogus, arguing both sides take 'fit in human spaces' too literally—it's not about physically squeezing into human-built houses. The real argument is safety.
- Safety for humanoids is wildly understudied, but he believes humanoids will ultimately be safer than mobile bases: they can be compliant in ways a mobile base cannot, working at higher rates in closer proximity to people at lower risk.
- The trade-off: mobile bases are safer in catastrophic failure (statically stable, can e-brake) while humanoids fall; but in the far more common case of bumping into people at close range, legs are safer—dynamic stability to catch yourself, and small actuators can control the CoM with exponential acceleration, whereas wheels would need giant dangerous motors and batteries.
- Industrial robots could already do your dishes, but you wouldn't be allowed in the kitchen with them; mobile bases are arguably less safe due to huge reflected inertia.
- Bottom line: the humanoid value case rests on compliance and near-human safety, not literal physical fit.
Related event: Engineer breaks down the 'humanoids fit human spaces' argument(2 posts)→
More from Embodied
- Robots will proliferate far faster than new homes get built, argues roboticist — KyleMorgenstein · 2026-11-20
- Codex-branded Raspberry Pi drives open-source AgOpenGPS tractor autosteer in the field — 4310sy · 2026-11-20
- PhD Researcher: 7 Years of Thesis Work on Humanoid Robots Fitting Human Spaces — KyleMorgenstein · 2026-11-20
- Legs vs wheels: the safety trade-off in robot form factors — KyleMorgenstein · 2026-11-20
- Same thread on humanoid safety — KyleMorgenstein · 2026-11-20
- NVIDIA Cosmos world models hit 20 million downloads on Hugging Face — heghbalz · 2026-11-20