The cycle that began with [YOU] on AI documents the twenty-fold productivity multiplier in Trivandrum, the solo founder shipping a revenue-generating product without writing a line of code, the engineer who built a complete user-facing feature in two days without prior frontend experience. Keynes predicted all of this in 1930. He also predicted that the response would be leisure—and the cycle's most uncomfortable evidence is that the prediction of material abundance was correct and the prediction about leisure was wrong. The builders do not stop. They post at three in the morning. The spouses write public letters asking for help. The tool that should have freed them has amplified the compulsion that Keynes identified as the permanent problem: the inability of human beings, freed from necessity, to know what they are for.
The cycle's analysis of this compulsion—what it identifies as productive addiction—is a case study in the dependence effect analyzed from a Keynesian angle. Keynes understood that work is not merely a cost. It is the primary source of identity, status, social connection, and psychological structure for the vast majority of people in industrial and post-industrial economies. To work is to be someone. To stop working is to face the permanent problem directly. Most people, confronted with that question, discover that they would rather produce. The AI tool did not create this psychology. It amplified it—at the precise historical moment when the economic conditions for addressing it are finally met.
The Keynesian diagnosis of the AI employment transition is structured around three failures of the classical assumption that supply creates its own demand. Speed: previous technological transitions displaced workers over decades; AI operates over months, and the gap between displacement and absorption is the space in which involuntary unemployment persists. Scope: previous automation targeted routine tasks; AI performs competently in domains previously considered non-routine, removing the ladder rungs that workers would have climbed to escape displacement. And the aggregate demand mechanism itself: if the productivity gains flow to capital and to the workers who direct AI tools while the displaced lose income, the aggregate demand that sustains the broader economy weakens in a contractionary spiral that individually rational headcount reductions produce collectively.
The paradox of thrift finds its AI equivalent in the attention economy: when every knowledge worker optimizes their cognitive time with AI tools, the aggregate result is not more genuine thought but less. The individually sensible decision to fill every pause with productive AI interaction depletes the attentional slack in which background processing—the integrative cognition that produces genuine insight rather than plausible text—operates. The fallacy of composition operates in the currency of attention: what is individually rational produces collectively impoverished outcomes.
John Maynard Keynes was born in Cambridge in 1883, the son of an economist and logician, and educated at Eton and King's College, Cambridge, where he returned as a fellow after a brief career at the India Office. He was a central figure in the Bloomsbury Group, a friend of Virginia Woolf and Lytton Strachey, a speculator who rebuilt King's College's finances after depleting them, and a civil servant whose judgment shaped British economic policy from World War I through the Bretton Woods conference of 1944, where he led the British delegation that created the postwar international monetary order.
The intellectual arc of his career moved from probability theory—his 1921 Treatise on Probability introduced the epistemological foundations he would later apply to economics—through the critique of British economic policy after World War I (The Economic Consequences of the Peace, 1919) to the two works that defined his legacy. A Treatise on Money (1930) laid the technical groundwork. The General Theory (1936) demolished the classical model that markets automatically tend toward full employment and prescribed the institutional response—fiscal policy, investment in human development, aggregate demand management—that governments have been implementing and arguing about ever since.
Keynes died in 1946, before the postwar prosperity that his prescriptions helped create. Robert Skidelsky's three-volume biography documents how the Keynesian synthesis was absorbed, diluted, and eventually repudiated by the neoclassical synthesis before being recovered, in modified forms, by the crises of 2008 and 2020. The AI transition presents a test case for a dimension of Keynesian economics that neither the synthesis nor its critics fully engaged: the institutional management of a transition so fast and so comprehensive that the mechanisms Keynes described for managing demand during displacement may be structurally insufficient.
Animal Spirits. Investment decisions cannot be made by rational calculation because the future is genuinely uncertain—not risky in the sense that probabilities can be assigned, but uncertain in the sense that the relevant probabilities cannot be calculated at all. What fills the void is animal spirits: the spontaneous urge to action, the gut-level conviction that this bet is worth making. Animal spirits are the engine of economic dynamism and the engine of speculative bubbles. AI tools have removed the friction that once filtered enterprise from speculation, liberating the spirits at precisely the moment when institutional moderation is least developed.
The Paradox of Thrift. When every household saves more, the aggregate result is less saving. The individually rational decision to save reduces spending, which reduces income, which reduces the capacity to save. The paradox of thrift applied to attention: when every knowledge worker optimizes their cognitive time, the aggregate result is less genuine thought—because the attentional slack in which background processing operates is depleted by exactly the optimization that felt productive at the individual level.
The Liquidity Trap of Capability. In 1932, monetary policy had done everything the textbooks prescribed: rates at historic lows, money available at minimal cost. Nothing happened, because the bottleneck had migrated from the supply of money to the demand for output. The capability trap is the AI-era analog: an economy saturated with productive capability that cannot convert the capability into value because the bottleneck has migrated from execution to judgment. More AI subscriptions do not escape the trap. Only judgment adequate to direct the capability does.
Sticky Identities. Keynes demonstrated that wages do not fall smoothly to clear the labor market because wage cuts carry meaning beyond their arithmetic: they are experienced as judgments about the worker's worth. Professional identity is stickier than any wage by an order of magnitude. The Python developer told to become a “creative director” is not being asked to learn a new skill. The developer is being asked to emigrate—to dissolve an architecture built through years of investment and construct a new one in a landscape where the markers of competence are unfamiliar and the community of practice that provided belonging has dissolved.
In the Long Run. “In the long run we are all dead” is not a bon mot. It is a methodological manifesto: economic theory must serve the generation that is living through the transition, not the hypothetical future generation that will benefit from the institutions eventually constructed. The short-run moral priority demands that the institutional response to AI displacement be designed for the people who are displaced now, not for the grandchildren who will benefit from the institutions built after the current generation has absorbed the costs.