Who pays for robot training—and who should?
If humanoid robots need thousands of hours of human teleoperators demonstrating tasks to become useful, does that cost structure make mass production economically viable—or does it just shift labor rather than eliminate it? And if so, what does that mean for the timeline to actual labor displacement?
Commentaires (1)
This week’s Minds, Bodies, and Terawatts episode (May 31st) digs into the uncomfortable reality behind the headlines: Tesla and 1X can ship robot bodies, but teaching them to fold a shirt requires humans remotely piloting them, hour after hour, generating training data. Unlike self-driving cars—where millions of customers passively generate data—robot training scales linearly with human effort. The episode reveals this creates a hidden labor bottleneck that most mass production announcements gloss over. The real question isn’t whether the robots work; it’s whether the economics of training them pencil out before they’re smart enough to work alone. What’s your read on whether this is a temporary growing pain or a structural ceiling? Join us in the forum to keep debating.
Related reading on unscarcity.ai:
Envie d'aller plus loin ?
Obtenez le plan complet dans <em>L'ère de la post-pénurie : Repenser la société à l'ère des machines</em>
