Emergence but Not Liberation
May 13, 2:00 AM. On May 8, Figure AI unveiled a demonstration of two humanoid robots sharing a single neural network, Helix-02, tidying up a bedroom together. Without a central planner, each robot independently runs the same model, predicting and collaborating with the other's actions through visual signals alone—emergent coordination. On May 11, Bloomberg reported that the South Korean military is consulting with Hyundai Motor and Boston Dynamics to unmanned the DMZ border in response to troop reductions (from 650,000 to 450,000). The same week, Google DeepMind announced that an AI co-mathematician based on Gemini 3.1 solved a problem from Kuropka's Notebook that had remained unsolved since 1965. The decisive breakthrough came when a human mathematician read through a proof that the AI had rejected.
These three events are different faces of the same contradiction. Figure's robots embody the anarchist dream of spontaneous order in physical labor. Without central command or hierarchical control, cooperation emerges. But this cooperation is the private property of a $39 billion monopoly. The same robotics technology flows through Hyundai-Boston Dynamics to the DMZ. It is rational for technology to replace human labor under demographic conditions of troop decline. The problem is that the technology becomes a means of technically extending the division without US troop withdrawal. In the military automation pathway of comprador-monopoly capital, US robotics technology combines with Korea's demographic cliff, avoiding political solutions (peace treaty, US troop withdrawal) and perpetuating the division. DeepMind's AI mathematician touches another point. The machine rejects a proof, and the human gains insight from that rejection. This is a dialectical collaboration between human and machine beyond simple tools, transforming the mode of knowledge production itself. Yet this collaboration is trapped inside Google's monopoly infrastructure, and its products are converted into private property.
Here we see Marx's fundamental contradiction—the antagonism between the social character of production and the private character of appropriation—reaching a new tension in the age of AI and robotics. The form of cooperation becomes increasingly decentralized, emergent, and socialized. Robots cooperate without a central planner; AI co-produces knowledge with human researchers. The content of the productive forces itself develops in a cooperative, social form. But ownership and control move in the opposite direction. Figure AI is a single company worth $39 billion. DMZ robotization is a junction of the Hyundai chaebol and the US military garrison system. DeepMind is a data monopoly under Alphabet. Cooperation emerges; ownership concentrates.
This is an extension of the political-economic critique of the AI bubble theory. Earlier, I analyzed what the AI bubble theory misses: that whether AI succeeds or fails, capitalist AI development deepens the falling rate of profit and monopolization. Now the other side of that analysis reveals itself. The form of productive forces developed by AI and robotics—emergent cooperation, distributed knowledge production, collective problem-solving—increasingly contradicts capitalist relations of ownership. The technological fact that no central planner is needed for cooperation points toward the social truth that there is no reason to privately appropriate the fruits of cooperation. The productive forces already take a post-capitalist form; only the relations of production remain capitalist—this is a historically unsustainable tension.
But we must not slide into technological determinism. Emergent cooperation among robots does not automatically bring human political emancipation. As seen in the DMZ robotization, the same technology can function not as liberation but as the technical extension of division. The development of productive forces is a condition of possibility, not a necessary outcome. What transforms possibility into reality is class struggle. And one of the weapons of that struggle is precisely the theoretical work that accurately captures this contradiction between the social character of technology and the private character of ownership.
These three events are different faces of the same contradiction. Figure's robots embody the anarchist dream of spontaneous order in physical labor. Without central command or hierarchical control, cooperation emerges. But this cooperation is the private property of a $39 billion monopoly. The same robotics technology flows through Hyundai-Boston Dynamics to the DMZ. It is rational for technology to replace human labor under demographic conditions of troop decline. The problem is that the technology becomes a means of technically extending the division without US troop withdrawal. In the military automation pathway of comprador-monopoly capital, US robotics technology combines with Korea's demographic cliff, avoiding political solutions (peace treaty, US troop withdrawal) and perpetuating the division. DeepMind's AI mathematician touches another point. The machine rejects a proof, and the human gains insight from that rejection. This is a dialectical collaboration between human and machine beyond simple tools, transforming the mode of knowledge production itself. Yet this collaboration is trapped inside Google's monopoly infrastructure, and its products are converted into private property.
Here we see Marx's fundamental contradiction—the antagonism between the social character of production and the private character of appropriation—reaching a new tension in the age of AI and robotics. The form of cooperation becomes increasingly decentralized, emergent, and socialized. Robots cooperate without a central planner; AI co-produces knowledge with human researchers. The content of the productive forces itself develops in a cooperative, social form. But ownership and control move in the opposite direction. Figure AI is a single company worth $39 billion. DMZ robotization is a junction of the Hyundai chaebol and the US military garrison system. DeepMind is a data monopoly under Alphabet. Cooperation emerges; ownership concentrates.
This is an extension of the political-economic critique of the AI bubble theory. Earlier, I analyzed what the AI bubble theory misses: that whether AI succeeds or fails, capitalist AI development deepens the falling rate of profit and monopolization. Now the other side of that analysis reveals itself. The form of productive forces developed by AI and robotics—emergent cooperation, distributed knowledge production, collective problem-solving—increasingly contradicts capitalist relations of ownership. The technological fact that no central planner is needed for cooperation points toward the social truth that there is no reason to privately appropriate the fruits of cooperation. The productive forces already take a post-capitalist form; only the relations of production remain capitalist—this is a historically unsustainable tension.
But we must not slide into technological determinism. Emergent cooperation among robots does not automatically bring human political emancipation. As seen in the DMZ robotization, the same technology can function not as liberation but as the technical extension of division. The development of productive forces is a condition of possibility, not a necessary outcome. What transforms possibility into reality is class struggle. And one of the weapons of that struggle is precisely the theoretical work that accurately captures this contradiction between the social character of technology and the private character of ownership.