Fauxtomation
Fauxtomation names the practice of disguising labor redistribution as labor elimination. Taylor coined the term in her 2018 Logic Magazine essay 'The Automation Charade' to describe technologies that transfer work from paid employees to unpaid users or low-paid invisible workers rather than eliminating work altogether. The self-checkout kiosk does not automate grocery scanning—it transfers that labor from a paid cashier to the customer, who performs it for free. The 'automated' content moderation system does not eliminate human review—it handles routine cases algorithmically and routes the most disturbing content to contract workers in the Philippines and Kenya earning poverty wages. The automation is fake in the sense that the human labor persists, but the fakery serves real economic functions: reducing visible labor costs, rendering remaining labor invisible, and maintaining the narrative that machines operate autonomously. Fauxtomation is the twenty-first-century version of the mechanism Cowan documented in the twentieth: labor-saving technology that saves the employer labor by transferring it to people with less power to refuse.
In the AI Story
Taylor's concept built explicitly on Cowan's foundation, extending the domestic paradox into the digital economy. The essay opens with direct citation of More Work for Mother and applies Cowan's analytical lens to contemporary platforms. The structural parallel is exact: the washing machine eliminated the laundress and transferred her labor to the unpaid housewife; the self-checkout kiosk eliminates the cashier and transfers their labor to the unpaid customer. In both cases, the technology works, the visible labor cost declines, and the actual labor is redistributed rather than eliminated. The redistribution follows existing power gradients: toward the people already designated as unpaid laborers (housewives, customers) or toward the people with the least bargaining power in the global labor market (data workers in the Global South).
The fauxtomation of AI systems operates at two levels. The first is the level Taylor identified: human labor essential to the system's operation is hidden from users to maintain the appearance of autonomous machine intelligence. The data labelers who annotate training data, the content moderators who filter outputs, the RLHF workers who refine model behavior—this labor force numbers in the millions globally and is structurally invisible in the user experience. The invisibility is functional; it preserves the narrative that AI 'just works,' which is essential to the technology's market value. The second level is novel to AI: the transfer of evaluation and quality-assurance labor to the end user. When a developer uses AI to generate code, the developer absorbs the labor of verifying correctness, maintaining consistency, and ensuring the output meets standards—labor that is invisible in the organization's productivity metrics but substantial in the developer's experienced workload.
Gray and Suri's Ghost Work provided the empirical documentation of fauxtomation's global labor infrastructure, revealing the millions of workers performing microtasks that make AI possible: labeling images, evaluating outputs, moderating content, correcting errors. These workers are paid piece rates that, like the domestic laundress's wages, are meager but real—and like the laundress, they are positioned for elimination as AI systems become more capable. The economic logic is identical: why pay humans when machines can do it? The question ignores that machines require extensive human labor to appear autonomous, but the labor is hidden well enough that the question feels reasonable to the people asking it.
Origin
Taylor developed fauxtomation during her research into the platform economy's labor practices, recognizing that automation was frequently a rhetorical claim covering labor redistribution. The term gained traction in technology criticism and labor organizing because it named a pervasive experience: the sense that 'automated' systems required more user effort, not less, and that the convenience advertised was often convenience for the platform rather than the user. The concept provided vocabulary for what users felt but could not articulate: that they were doing work that employees used to do, for free, while being told the system had become more efficient.
Key Ideas
Automation rhetoric conceals redistribution. Technologies marketed as autonomous often depend on invisible human labor transferred to users or to hidden workforces—the appearance of automation serves economic functions regardless of technical reality.
The customer becomes the unpaid worker. Self-service technologies transfer labor from paid employees to unpaid customers under the euphemism of convenience—a redistribution that benefits the platform while increasing the user's total effort.
Global labor arbitrage enables the illusion. AI systems appear autonomous partly because the human labor sustaining them is performed by workers in countries where wages are low enough that the labor costs less than the marketing budget obscuring it.
Fauxtomation is the Cowan paradox accelerated. What took decades in the domestic sphere—the gradual elimination of paid domestic workers and transfer of their labor to unpaid housewives—is happening in years in the digital economy, at global scale, with platforms as intermediaries.
Appears in the Orange Pill Cycle
Further reading
- Astra Taylor, 'The Automation Charade,' Logic Magazine 5 (August 2018)
- Mary L. Gray and Siddharth Suri, Ghost Work (Houghton Mifflin Harcourt, 2019)
- Ruth Schwartz Cowan, More Work for Mother (Basic Books, 1983)
- Lilly Irani, 'Difference and Dependence among Digital Workers,' South Atlantic Quarterly 114, no. 1 (2015)