Most prognosticators describe the flow like this: AI performs white-collar work, then, embedded in robots, performs blue-collar work — and at that point society suffers questions around power, purpose, and participation.
Broadly that’s right. But it needs one correction: white-collar first, blue-collar once robotics matures describes which labor markets get disrupted, not which contract of standing is being broken. And the contracts are the real story.
The achievement contract breaks first
Modern society organized itself around a powerful bargain — the achievement contract:
Develop competence → perform economically useful work → earn resources, status, and autonomy → participate in society.
AI attacks the middle of that equation. If increasingly capable systems can write software, analyze contracts, diagnose problems, design products, produce media, do research, and manage logistics, then competence no longer guarantees economic scarcity. A person can spend twenty years becoming excellent at something and find the field no longer needs much human labor to do it.
That is not merely job displacement. It challenges the proposition that your differentiated achievement is the basis of your standing — which is why the coming crisis is much larger than unemployment.
When robotics arrives, the role contract breaks too
Blue-collar automation eventually reaches something older than achievement: the role contract. Do your duty. Fill your role. Contribute. Be dependable. Earn your place within the larger order.
For enormous numbers of people, work provides exactly that. “I am a mechanic.” “I am a nurse.” “I build houses.” These aren’t income descriptions; they are locations inside a moral and social structure. If machines increasingly perform the role, the question becomes not what will I earn? but what am I for?
So there are two waves. Knowledge automation destabilizes achievement: what happens when expertise is no longer scarce enough to organize status and income? Robotics destabilizes role: what happens when socially necessary labor no longer requires everyone to occupy a productive position?
The legitimacy crisis
Together, those waves land on exactly three questions: who has power, what gives life purpose, and on what basis does someone participate? These push toward a response that insists human standing cannot depend solely on economic productivity — opening questions of universal provision, dignity independent of employment, shared ownership, and access.
Underneath it is a decoupling. The inherited equation —
labor → income → legitimacy → participation
— is revealed to be not ontologically necessary. That recognition is the easy part. The hard part is that saying “everyone deserves participation” does not tell you how to organize a civilization in which production, political power, meaning, ownership, technological capacity, and human participation have come partially uncoupled.
Why it becomes a systems problem
No single moral principle solves that. To design the replacement you have to model, at once: economic production, capital ownership, AI access, energy, taxation, education, identity, status, leisure, psychological development, governance, inequality, family formation, community, technological concentration, and meaning. That is not an ethics question with a slogan for an answer. It is a systems problem — the kind that needs whole-system modeling rather than a single fix.
Power is the dangerous hinge
Here is the twist: AI does not automatically produce the humane outcome. The opposite is entirely possible. Suppose a small number of corporations or governments control frontier AI, robots, energy, compute, capital, data, and manufacturing. Then most labor becomes less necessary while ownership stays highly concentrated.
That is not a fairer society. It is closer to an extraordinarily amplified concentration of raw power: those who control the machines gain unprecedented leverage over everyone who no longer controls economically necessary labor. The result would be one of the strangest configurations imaginable — civilization-scale technological capability sitting on top of old institutions and a concentration of power, while populations demand a participation the structure was never redrawn to give. That is unstable by construction. Whether these capabilities are enclosed or held as a commons is the fork that decides which way it tips.
The participation question
So the deepest crisis may not be what will people do for work? It may be:
On what basis does a human being have a commensurate claim upon the civilization when their labor is no longer required by it?
Today the answer is largely: you participate because you produce. AI potentially breaks that, and then a civilization has to discover another answer — a new basis for standing that does not route through economic necessity.
That is why the transition feels epochal. Industrialization changed what humans did for work. Advanced AI could challenge why work has been the principal mechanism through which people receive resources, status, purpose, and standing at all. The task is not to cushion unemployment. It is to construct a viable replacement ecology before the old contracts finish breaking.