This post expands on an article first presented at the Social Design Network Conference in Lugano in September 2025.

A couple of days before I presented my contribution at the conference in Lugano, Elon Musk’s friend and Salesforce CEO Marc Benioff posted a video of Tesla’s Optimus robot on X — with this caption:

Elon’s Tesla Optimus is here! Dawn of the physical Agentforce revolution, tackling human work for $200K–$500K. Productivity game-changer! Congrats @elonmusk, and thank you for always being so kind to me!

Two aspects of this post are striking. The first is the euphemistic framing. By claiming that Optimus will “tackle human work,” the post sidesteps the far more contentious implication of human replacement, while preserving its economic promise. The second is the stark disconnect between the celebratory rhetoric and the robot’s actual performance in the video. Futurism summarized the demo shown in the video with brutal clarity: "Elon Musk’s New Optimus Robot Demo Is So Painful It Will Make You Wince".

This mismatch is precisely the core of my argumentation. The claim that humanoid robots will soon replace human workers remains highly speculative, but the narrative surrounding that claim is already exerting real effects. The power of the humanoid robot lies less in its technical capabilities than in its symbolic force. By repeatedly invoking a future in which human labor is rendered obsolete, these narratives subtly erode the perceived value of work, weaken workers’ bargaining power, and normalize the dismantling of labor protections. Structural inequalities are reframed not as political choices, but as the unavoidable outcome of technological progress, benefiting those who profit from automation while avoiding responsibility for its social costs.

Recently, a new video of the Optimus robot made headlines. At an event in Miami, the robot pauses, raises its hands as if a human operator is removing a VR headset, and then falls over. Many observers saw this as a clear indication that, contrary to Elon Musk’s claims, the robot was not acting autonomously, but was being teleoperated by a human.

This example illustrates why the debate over humanoid robots cannot be reduced to questions of feasibility or performance—and why I submitted an article to the Social Design Network Conference specifically. The real issue lies in design and narrative. As a robotics researcher, I have observed a troubling shift in the narrative around humanoids since my Ph.D. in 2011. What was once a nerdy fantasy has become a discourse about human worth. When machines are deliberately made to look like us, they invite direct comparison, and in doing so, they quietly turn human workers into benchmarks to be beaten rather than contributors to be supported.

As I mentioned in the article, a similar logic is at work in the marketing strategies of AI startups such as Artisan. The company sells AI assistants branded as “Artisans,” explicitly framing them as virtual, human-like employees. Its advertising campaign is deliberately provocative. Slogans such as “Artisans won’t complain about work-life balance” and “Excited to work 70+ hours a week” portray human labor as inefficient, fragile, and inconvenient. The message escalates to an explicit call to “Stop hiring humans,” culminating in the slogan: “The Era of AI Employees Is Here.”

Yet even Artisan’s CEO acknowledges that this supposed era remains hypothetical. On the company’s own website, he concedes: “Eventually, we should live in a world where [...] you can truly stop hiring humans, but today is not that day”. The campaign is therefore not about promoting an existing, functional alternative to human labor. It is about promoting a vision of the future that aligns closely with the interests of the company and, more broadly, with a segment of the tech elite. The objective is to shape expectations, norms, and power relations long before the technology can justify them.

A substantial body of research shows that forcing automation where it does not align with human, organizational, or technological realities can be detrimental for companies, pointing instead toward collaborative human–machine approaches. As MIT economist David Autor and Google SVP James Manyika argue in a recent Atlantic article, “A Better Way to Think About AI: Artificial Intelligence Is Ready to Collaborate. Why Fixate on Automation?”, collaborative approaches that combine the complementary strengths of humans and machines deliver better performance than attempts at partial, low-quality automation driven by the hope that systems will eventually meet expectations and replace human work.

In their Harvard Businesse Review article "A smarter strategy for using robot", MIT automation experts Ben Armstrong and Julie Shah further show that,“companies are best served by a positive-sum automation that draws on the strengths of intelligent machines, managers, engineers, and line workers alike. The vision is not one without humans, but one in which automated systems make humans more capable and more vital at work.” However, realizing this vision requires a profound rethinking of the production worker: not as a liability or source of error, but as a central contributor to innovation, adaptability, and system resilience.

As pointed out by the historian of technology Martina Hessler in her book Sisyphos im Maschinenraum, our relationship with machines has been guided by the notion that they exist to transcend human limitations since the early nineteenth century. The image of a perfect machine that can solve every conceivable problem has since had a decisive influence on our approach to technological development. Challenging this framing is essential if we are to move toward more flexible and resilient work processes that leverages human strengths. It also means confronting the power structures and financial incentives embedded in dominant automation narratives. Success metrics need to be redefined away from narrow, short-term efficiency gains and toward long-term resilience, human capital development, and collective performance.

Revalorizing human work begins with a critical reassessment of the design choices behind humanoid robots. Design is never neutral. When robots are given human form, they invite direct comparison with human workers, a comparison that is rarely benign. Humanoid design reinforces the technocratic imaginary of the perfect machine, in which humans appear as inefficient precursors to their artificial successors.

The future of work is not predetermined by machines. It is actively constructed through the narratives we promote, the designs we legitimize, and the values we embed into our technologies. Recognizing this is the first step toward reclaiming technological progress as a collective project that works for society as a whole, not just for the big tech players.

The oft-cited phrase attributed to Henry Ford, “If I had asked people what they wanted, they would have said faster horses”, is usually invoked to justify disruptive innovation. In the case of humanoid robots, however, the metaphor cuts the other way. These machines are the faster horses of automation: familiar forms repackaged as progress, masking a vision of innovation driven less by progress and long-term value creation than by labor substitution and short-term cost efficiency.

The image of the Tesla Optimus Robot is licensed under the Creative Commons Attribution 3.0 Unported