CONCEPT
Self-Programmable Labor
Castells's term for the class of workers with the capacity to <em>retrain, adapt, and redirect</em> their skills — a capacity AI makes both more valuable and harder to acquire.
Self-programmable labor names the critical labor-market distinction of the network society: between workers with the capacity to retrain, adapt, and redirect their skills in response to changing technological conditions, and generic labor, whose work can be substituted by machines or by other generic workers anywhere in the global network. The distinction is not about current skill level but about the capacity to acquire new skills as conditions change. AI intensifies the distinction by raising the threshold of self-programmability — the base level of skill and adaptability required to remain in the self-programmable category. In the pre-AI information economy, self-programmability required the capacity to learn new tools, collaborate across contexts, and participate in knowledge networks. In the AI-augmented economy, it requires additionally the capacity to direct AI systems toward useful ends, evaluate their outputs against domain standards, and maintain the deep expertise necessary to recognize what the tools get wrong.
In The You On AI Field Guide
The self-programmable worker has existed throughout industrial history in various forms