Can robots become simultaneously capable, safe, inexpensive, and reliable in messy human environments?
Skeptics argue touch-rich manipulation lacks both the data and the forgiving physics that made language tractable. Capable, safe, cheap, and reliable is a four-way tradeoff no one has yet closed in an open environment.
View on the map → · Open in Browse →
The skeptics published their receipts: Brooks' eighth annual scorecard extends the prediction that deployable dexterity stays far below human hands beyond 2036, the 1X NEO launch conceded the point by shipping teleoperation, and the standards world named hands — grasp, adjustment, recovery — as the constraint humanoid programs keep routing around.
Despite billions in investment, startups say their androids mostly aren't useful for industrial or domestic work yet
Brooks' eighth annual scorecard extends his canonical dexterity skepticism into new ten-year predictions: deployable dexterity will remain far below human hands beyond 2036, and walking humanoids will stay unsafe near people.
Humanoids will not take off until we solve the hands...A humanoid that cannot grasp, adjust, feel, and recover is not a helper. It is a prop.
ASTM's robotics standards director argues manipulation — grasp, grip adjustment, error recovery on soft and irregular objects — is the binding constraint, from inside the standards and testing world.
I've yet to see a humanoid complete a household chore from start to finish.
The reference piece in the 1X NEO teleoperation controversy: no humanoid can yet finish a real household chore autonomously, so home robots will launch as remote-worker embodiments — a concrete measure of the capability-reliability gap in messy environments.