Jean-Baptiste Say's 1803 Traité d'économie politique argued that production and consumption are reciprocal acts linked through exchange. When a farmer grows wheat, the act of growing it generates wages, rent, and profit — income that constitutes demand for other products. The circuit is self-reinforcing: production generates income, income constitutes demand, demand motivates further production. This is the actual Say's Law: a claim about the aggregate relationship between production and income in functioning market economies, not the caricature that every individual product automatically finds buyers. Understanding the circuit — its operation, its disruption, its re-establishment — is essential to reading the AI transition as an economic phenomenon rather than merely a technological one.
The circuit's precision is what makes it analytically powerful. Say was not claiming that overproduction of specific goods was impossible — he was a cloth manufacturer who had witnessed business failure directly. He was claiming that general gluts — situations where total production exceeds total demand across the entire economy — are structurally impossible, because total production and total demand are two sides of the same transaction. The income generated by producing goods is the purchasing power that constitutes demand for other goods. You cannot, in aggregate, produce more than the economy can absorb, because the producing itself is what creates the capacity to absorb.
The AI economy tests the circuit in a way Say could not have anticipated. Production is expanding dramatically — a single person with a natural language interface can build what once required a team and a year. But the income side of the equation is being rerouted through a much narrower channel. If five engineers can do the work of a hundred, the income that ninety-five engineers would have generated does not materialize. The circuit is disrupted not at the point of production but at the point of income distribution — a possibility Say's framework identifies but that his simplifiers typically ignore.
Say's response to David Ricardo's chapter on machinery is instructive. When Ricardo reversed his earlier position and admitted that machinery could be detrimental to the laboring class, Say did not deny the displacement. His argument was more specific: that the market would adjust, that displaced workers would find employment in new sectors created by the very productivity that had displaced them. The key word is adjusted. Say understood adjustment takes time, that transitional costs fall on real people, and that his structural optimism did not retroactively justify the suffering of those caught in the middle.
The circuit remains the most precise instrument available for analyzing the AI transition because it identifies both the mechanism and its vulnerability. Production expanding while the income circuit constricts is exactly the configuration Say's framework treats as a transitional challenge — real, consequential, and eventually self-correcting through the emergence of new production that generates new income. The question is not whether the circuit will re-establish itself, but how long the adjustment will take and who will bear the cost.
Say published the Traité in 1803, two years after Napoleon suppressed a tribunate he had served in. The book was intended as a systematic popularization of Adam Smith's Wealth of Nations, reorganized and corrected where Say believed Smith had erred. Its success was enormous: it went through multiple editions, was translated widely, and established Say as the leading French economist of his generation. The circuit argument was the book's central theoretical contribution.
The simplification began almost immediately. By the 1820s, the phrase 'supply creates its own demand' — never written by Say in that form — was circulating as the canonical statement of his position. John Maynard Keynes's 1936 attack on 'Say's Law' crystallized the simplified version as the canonical adversary, and generations of economists on both sides have argued about the caricature rather than the original.
Reciprocity, not automatic clearance. The circuit links production and demand through the medium of income, but does not guarantee that individual transactions will always clear or that the circuit will operate at the same speed in every circumstance.
Aggregate, not individual. Say's claim applies to the economy as a whole, not to specific products. Individual goods may fail to find buyers; the economy as a whole cannot suffer a general glut.
Conditional on adjustment time. Say's structural optimism is paired with an implicit acknowledgment that adjustment takes time and that transitional costs are real and unevenly distributed.
Diagnostic for disruption. The circuit's vulnerability — the possibility that production can expand while the income circuit contracts — is the specific analytical tool that makes Say's framework applicable to the AI transition.
The simplified version of Say's Law remains the dominant reception in both defenses and critiques. Defenders continue to invoke 'supply creates its own demand' as grounds for laissez-faire indifference to transitional costs. Keynesian critics continue to attack the same caricature as evidence that state intervention is necessary. The original framework, which treats the circuit as real but vulnerable, remains largely unused in contemporary economic argument despite being more analytically powerful than either position.