Amid the hype surrounding humanoid robots, a sober voice from investors is being heard more and more often, warning that the market may be entering an overheating phase. This is not about the technology “not working”—on the contrary, progress in sensing, control, manipulation, and embodied AI is obvious. The problem lies elsewhere: economics is not yet keeping pace with the promises.
Most humanoid startups today showcase impressive prototypes, but remain very far from scalable business models. Development and production costs are high, maintenance is complex, reliability in real-world conditions is limited, and a clear ROI exists only in narrow niches—logistics, pilot industrial operations, security. The mass market and the “universal robot” remain, for now, more of a narrative than a commercial reality.
Investors are wary of a familiar pattern: money is chasing a vision of the future rather than proven demand. Company valuations are growing faster than the number of real, working deployments. This is a dangerous asymmetry, already seen in previous technological waves—from the dot-com era to certain segments of Web3.
It is important to note that this is not about the collapse of the field, but about a possible correction. Humanoid robotics will almost certainly remain part of the future economy, but the path to it will be longer, more expensive, and more boring than slide decks suggest. The winners will not be those who speak the loudest about “robots everywhere,” but those who can prove effectiveness in concrete, down-to-earth scenarios.
The investors’ warning is not a signal to flee, but a signal for the market to mature. If expectations are not brought into line with reality, the current boom could easily turn into a sobering lesson in technological restraint.
