
The [YOU] on AI cycle documents the exhaustion of possibility in its most candid passages: the author lying awake at night, unable to turn off the part of his brain that kept optimizing; the grinding compulsion that replaces exhilaration; the engineer who spent the first two days of the Trivandrum training oscillating between excitement and terror. Bröckling’s framework identifies what those confessional moments are describing at the structural level: not personal weakness but the predictable consequence of a governing rationality that has no concept of enough encountering a tool that has no capacity to say stop.
The cycle uses the exhaustion of possibility to complicate its own optimism about what AI democratizes. The Berkeley researchers documented workers expanding into new domains not because anyone asked them to but because the tool made expansion possible and the internalized imperative made it feel necessary. Designers started writing code. Managers started building prototypes. Each expansion is a genuine gain in capability—the floor of what any individual can do has risen. Each expansion is also a new domain requiring optimization, a new arena in which the permanent tribunal will evaluate performance. The democratization of building democratizes the obligation to build. When crossing a boundary costs nothing, the entrepreneurial self crosses every boundary, because not crossing is experienced as failure to optimize.
The exhaustion of possibility is the cycle’s answer to its own most pressing question—the flow-compulsion problem. Bröckling identifies the structural limitation that the individual-level distinction cannot resolve: the regime that produces the auto-exploiting subject does not merely teach her to exploit herself. It teaches her to experience the exploitation as self-realization—to feel the compulsion as flow, the optimization as passion. The exhaustion of possibility is the state at the far end of the optimization cycle, when the compulsion has been experienced as flow for long enough that the exhilaration has drained and what remains is the grinding that the cycle names but cannot, from within its own frame, fully diagnose.
Bröckling develops the concept at the intersection of three earlier analyses. From Ehrenberg, the sociology of depression as the pathology of the imperative society: when prohibition is replaced by injunction (“you can, you must, you should”), the characteristic psychopathology is not neurosis but the depressive exhaustion of a subject who faces infinite demand with finite adaptive resources. From Han, the phenomenology of the burnout society: the burnout society’s achievement subject who exploits herself while experiencing the exploitation as freedom. From his own framework, the institutional history of how the entrepreneurial self was manufactured through specific governing technologies—the performance review, the coaching session, the creativity imperative—each of which contributed to the internal governance apparatus that now operates at machine speed.
The specific AI application of the concept emerged from Bröckling’s reading of the empirical literature on AI at work, including the Berkeley study that documented workers who expanded into adjacent domains, filled every pause with productive activity, and reported burnout alongside productivity gains. The burnout and the productivity were the same phenomenon: the AI having removed the execution constraint, the entrepreneurial self experienced both the freedom and the demand as unlimited, and the unlimited demand, meeting a finite biological system, produced the specific clinical syndrome that Ehrenberg’s framework had predicted thirty years earlier for a society organized around the imperative to act.
Infinite possibility as infinite demand. The collapse of the imagination-to-artifact ratio does not liberate the entrepreneurial self from the optimization imperative. It removes the last constraint on that imperative’s expression. When every idea can be realized, every unrealized idea is an optimization forgone, and the permanent tribunal records the deficit. The subject does not experience more freedom; she experiences more demand disguised as freedom.
The criterion problem. In a world of bounded possibility, the criterion for choosing among possibilities is partly provided by the bounds themselves: you cannot pursue everything, so the question of which things to pursue is answered partly by which things are possible. When the bounds dissolve, the criterion must come from somewhere else—from values, from judgment, from the kind of substantive theory of the good that the entrepreneurial regime treats as irrelevant and the Confucian tradition treats as the central question. The exhaustion of possibility is, in part, an exhaustion of the criterion.
The biological asymmetry. AI compresses the optimization cycle from weeks to minutes. The body does not compress accordingly. Integration—the slow process through which experience is converted into understanding, through which information in working memory is consolidated into long-term knowledge, through which the pauses and the friction and the resistance of the world build the kind of embodied expertise that ascending friction theory locates at every higher cognitive floor—requires the time the optimization imperative eliminates. The exhaustion of possibility is also the exhaustion of the biological system that must integrate at biological speed what the tool produces at machine speed.
The phenomenological trap. The exhaustion of possibility is especially difficult to recognize because the regime has taught the subject to experience compulsion as passion. The distinction between flow and compulsion loses its diagnostic reliability when both produce the same subjective texture from the inside. The exhaustion of possibility becomes legible only from outside the engagement, in the retrospective recognition of the hollow quality that distinguishes compulsive production from genuine flow—tired and hollow rather than tired and full, as [YOU] on AI’s epilogue puts it.