The cycle uses Taylor as the primary analytical lens for what is at stake in the organizational response to AI. The question is not whether AI multiplies productivity—it does, by a factor of five to twenty across a growing class of knowledge work—but who captures the gain and in whose interest the tool is deployed. Taylor believed, with genuine conviction, that scientific management served workers and employers alike: the worker who loaded more pig iron per day earned more money per day. The harmony never materialized. The gains went disproportionately to capital; the costs fell disproportionately on labor; the one best way turned out to be the best way for the owner, not necessarily for the owned. The AI transition faces the identical distributional question, and the cycle insists on naming it.
The cycle identifies two responses to the twenty-fold productivity multiplier. The organization that converts it into a twenty-fold headcount reduction captures the gain as margin and applies Taylor’s logic to a new medium: the stopwatch has become the algorithm, the one best way is still imposed from above, and the worker is still reduced to a system. The organization that deploys AI to amplify its workers—giving each person the capability to operate across domains, to direct execution rather than perform it—is doing something Taylor never conceived: treating the worker not as a component to be optimized but as a mind to be empowered. The choice between these responses is not technological. It is moral.
The cycle’s most direct engagement with Taylor concerns decomposition. Taylor decomposed work to eliminate the need for skilled workers; the skilled machinist who set his own cutting speeds was replaced by a sequence of specialists executing one fragment each, with the knowledge of the whole extracted into instruction cards held by management. Software development organized itself on exactly these principles—the product manager, the designer, the frontend developer, the backend developer, the QA engineer, the DevOps engineer, each performing a fragment that no single person was expected to comprehend. AI has collapsed that decomposition not by improving each fragment but by enabling a single person to perform the whole. The fragments were never intrinsic to the work; they were a workaround for the limitation that AI has removed.
Frederick Winslow Taylor was born in 1856 in Philadelphia into a Quaker family of comfortable means. He declined to attend Harvard, worked as an apprentice pattern-maker and machinist at the Enterprise Hydraulic Works, and in 1878 took a position at Midvale Steel, where he spent twelve years developing the methods that would make him famous. He began as a common laborer and rose to chief engineer, conducting the time-and-motion studies that would eventually comprise the core of scientific management. His investigations were not merely observational; he was deeply interventionist, redesigning workflows, adjusting incentive structures, timing every motion with a stopwatch, and extracting from the bodies of skilled workers the embodied knowledge that had previously resided only in those workers.
Taylor’s most famous experiment, conducted in 1899 at the Bethlehem Steel Company, concerned a worker named Henry Noll, whom Taylor called “Schmidt,” who loaded pig iron. Under Taylor’s redesigned method Noll’s output rose substantially, and his wages rose accordingly—but the number of men required to perform the same total work fell dramatically. The Principles of Scientific Management, published in 1911, codified these methods and argued for their universal application. In 1912, Taylor testified before a special committee of the House of Representatives investigating whether scientific management was, as its critics charged, a system designed to speed up workers until they broke. Taylor was characteristically certain that it was not. The workers who had testified before him told a different story. Both stories were accurate about their respective domains.
Taylor died in 1915, at fifty-nine, having spent his last years in evangelism for scientific management. Peter Drucker called him the man who made possible “all of the economic and social gains of the twentieth century.” The claim is extravagant only if you have not studied the evidence. Taylor’s methods—decomposition, time-and-motion study, standardization, the separation of planning from execution—organized the industrial economy that built the modern world. They also organized the alienation that accompanied the building, and the question of whether the gains were worth the human costs is one that the AI age is forcing open again, in a new medium, with an urgency that Taylor himself would have found difficult to explain.
The one best way. For every task performed by a human being, there exists a single optimal method, discoverable through scientific observation of the task’s elementary operations, codifiable through instruction, and enforceable through training and incentive. The worker who chose her own method was not exercising autonomy. She was introducing inefficiency. The one best way must be imposed, for the good of both the employer and the employee. This axiom is now being inverted: the AI era’s one best way is a tool distributed to every individual in the organization, who then directs it according to her own understanding of what needs to be built and why.
Decomposition and the logic of specialization. Complex work must be broken into elementary operations, each simple enough to be analyzed, timed, standardized, and assigned to a worker who needs no understanding of the whole. Decomposition was never intrinsic to the work; it was a workaround for the limitation that no single person could handle the whole. Remove the limitation, and the decomposition becomes what it always was: overhead. Coordination costs. Handoff losses. The sixty percent of development time consumed by communication, meetings, documentation—all of it the cost of decomposition, none of it the cost of building.
The separation of thinking from doing. Management thinks. Workers execute. The planning department designs the process; the instruction card specifies the motions; the worker follows the card. This division is the deepest structural feature of Taylorist organization, and it persists in every org chart, every job description, every sprint cycle. AI collapses the division not because thinking has been automated but because the doing has been, which frees every worker for the thinking that Taylor reserved for a managerial elite.
The worker as system. “In the past, the man has been first; in the future the system must be first.” This prescription, offered by Taylor as a moral claim, installed the assumption that the proper relationship between human beings and their work is the relationship between a system and its components—and that assumption is now so deeply embedded in organizational culture that it has become invisible. The AI age makes it visible by making it optional. For the first time, organizations have a genuine choice: deploy AI to optimize the worker, or deploy AI to amplify the worker. The tool does not determine the choice. The assumptions do.
The objectivity of measurement. Taylor believed that scientific measurement removed the personal and political from the relationship between labor and management: the stopwatch did not have preferences, and therefore its verdict was trustworthy. Algorithmic management has inherited this faith wholesale: the algorithm is neutral, the data is objective, the score is simply a fact about the worker. The workers in Amazon’s fulfillment centers, governed by systems that specify their movements, measure their speed, and discipline their deviations, are working in the precise organizational structure Taylor designed—with the stopwatch replaced by a sensor network and the foreman replaced by a model.