AI "Luddite" Concern is About Money, Not Technology

Perhaps we should not be surprised that polls show young people are wary of artificial intelligence, fearing it will take jobs. Such distrust seems a recurring feature of economic and technology history.

Period / technology	Workers or groups affected	Nature of the distrust	What actually happened
1811–1816: mechanized textile looms	English handloom weavers, framework knitters, textile artisans	Luddites destroyed machinery they believed was being used to replace skilled labor and cut wages.	Many traditional textile skills really did decline. The textile industry became increasingly mechanized, although employment did not simply disappear; production and markets expanded. The original Luddite grievances were therefore partly about distribution of income and control over work, not just technology. (National Archives)
Early–mid 1800s: steam power and railroads	Canal workers, coachmen, agricultural workers, some local trades	Railroads were portrayed as dangerous to existing occupations and disruptive to communities and established economic arrangements.	Railroads destroyed or diminished some occupations while creating enormous new industries and occupations. Opposition also reflected land, environmental and social concerns—not simply employment. (DigitalCommons)
Late 1800s–early 1900s: industrial machinery	Skilled craftsmen, factory workers, agricultural laborers	Mechanization was feared to replace skilled human labor with machines and reduce workers to machine tenders.	Industrial productivity exploded. Many occupations disappeared or shrank, but manufacturing employment and entirely new industries expanded for long periods. The nature of work changed dramatically.
1920s–1940s: automatic telephone switching	Telephone operators, predominantly women	Mechanical switching threatened one of the most visible employment opportunities for women.	AT&T automated more than half of the U.S. telephone network between 1920 and 1940, eliminating most operator jobs. Yet research finds that employment among subsequent generations of young women was not reduced overall because new clerical and service occupations expanded. Incumbent operators, however, suffered significantly. (National Bureau of Economic Research)
1940s–1960s: farm mechanization	Agricultural laborers, farmhands	Tractors, combines and other machinery dramatically reduced the need for manual agricultural labor.	Agricultural employment collapsed as productivity soared. Workers moved into manufacturing, construction and services. This is one of the clearest historical examples of technology eliminating an enormous number of jobs while the economy simultaneously became much richer.
1950s–1960s: computers and industrial automation	Factory workers, railroad workers, clerical workers	This produced a remarkably modern debate about “technological unemployment.” Organized labor worried that automation would eliminate entire categories of work.	The concern was serious enough that the Kennedy administration created programs for retraining and adjustment. The U.S. Department of Labor records widespread fear that automation would produce mass permanent unemployment. (U.S. Department of Labor)
1960s: automation in manufacturing	Automobile, steel, railroad and coal workers	Workers saw machines producing more output with fewer people. Kennedy himself warned of “industrial dislocation” and unemployment.	Productivity increased dramatically while employment shifted toward other sectors. But the adjustment was painful and geographically concentrated. Kennedy noted that railroads, coal mines and steel mills could produce more with substantially fewer workers. (JFK Library)
1970s–1990s: ATMs	Bank tellers	It seemed almost inevitable that machines dispensing cash and accepting deposits would eliminate tellers.	Something more interesting happened. ATMs reduced the number of tellers needed per branch, but they also reduced the cost of opening branches. More branches and banking services partly offset the productivity effect. Tellers' work shifted toward customer service and sales. (IMF eLibrary)
1970s–2000s: computers in offices	Typists, secretaries, bookkeepers, clerical workers	Personal computers and office software appeared capable of replacing enormous amounts of routine administrative work.	Many particular occupations contracted sharply, but computers also created entirely new categories of employment. The transformation was nevertheless highly unequal: workers whose skills complemented computers benefited more than those whose tasks were automated.
1990s–2010s: Internet and e-commerce	Retail clerks, travel agents, newspaper workers, postal and publishing workers	The Internet appeared likely to eliminate intermediaries and many traditional information jobs.	Many traditional occupations shrank or changed dramatically. At the same time, software, logistics, digital advertising, e-commerce and online services created new economic activity.
2010s–2020s: AI and generative AI	Writers, programmers, customer-service workers, analysts, designers and other knowledge workers	Unlike earlier automation, AI threatens portions of cognitive and creative work, raising fears that even highly educated workers may become economically redundant.	The outcome is still unsettled. Evidence increasingly suggests that AI can substitute for particular tasks while also increasing productivity and creating new tasks. The important historical question is likely to be how quickly workers and institutions adjust, rather than simply whether AI eliminates jobs.

Ironically, though we habitually assume younger people are digital native and comfortable using technology, they also are not immune from logical concerns about how new technology will reshape job markets. 

If history applies to AI as it has in the past, job tasks and functions are very likely to be disrupted. Some jobs will disappear as well. 

But new jobs will be created. And younger people are likely to be the beneficiaries, compared to older workers. 

Effect	Historical precedent	Likely significance for AI
Some tasks disappear	Virtually every major technological revolution	Very high
Some occupations shrink substantially	Weavers, telephone operators, agricultural workers, typists	High
New occupations and industries emerge	Industrialization, computers, Internet	Very likely, but difficult to predict
Workers displaced today automatically benefit from new jobs tomorrow	Historically not necessarily	Older workers tend not to benefit

A 2026 study by David Autor, Caroline Chin, Anna Salomons and Bryan Seegmiller, for example found that new work is disproportionately performed by younger and more educated workers, even after controlling for occupation, industry and location. 

Study	Technology / setting	What it finds relevant to age	Implication
Autor, Chin, Salomons & Seegmiller (2026), “What Makes New Work Different from More Work?”	U.S. occupational change, 1940–2023	New work is disproportionately performed by younger and more educated workers. New work also carries significant wage premiums that are larger for newer occupations. (National Bureau of Economic Research)	Very strong support for the idea that technological/economic change creates opportunities disproportionately captured by younger workers. NBER study
Deng, Müller, Plümpe & Stegmaier (2024), “Robots, Occupations, and Worker Age”	German manufacturing plants adopting robots	Robot adoption did not reduce employment for an entire age group, but the reinstatement/creation effect was age-biased toward young workers. The authors conclude that young workers benefited most from the new jobs created by robot adoption. (IZA)	Perhaps the closest direct evidence to your proposition: displacement can be occupation-specific while the new employment generated by technology disproportionately favors younger workers. IZA study
Battisti, Dustmann & Schönberg (2023), “Technological and Organizational Change and the Careers of Workers”	German firms and technological/organizational change	Firms often retrain routine workers into more abstract jobs, but older workers are an important exception: technological/organizational change increases their risk of permanently leaving employment and reduces earnings. (IZA)	Strong evidence for an age asymmetry in adjustment: younger/mid-career workers can be retrained into new work more successfully. IZA study
Kogan, Papanikolaou, Schmidt & Seegmiller (2023/2025), “Technology and Labor Displacement”	U.S. patents matched to worker-level administrative data	Labor-saving technologies reduce exposed workers' earnings. Labor-augmenting technologies increase employment, but earnings gains are concentrated among new entrants, while earnings can decline among incumbents—especially older, white-collar and higher-paid workers. (National Bureau of Economic Research)	Very strong evidence for a newcomer-versus-incumbent effect. Technology can create opportunities while simultaneously eroding the value of incumbent expertise. NBER study
Kogan et al. (2021/22), “Technology, Vintage-Specific Human Capital, and Labor Displacement”	U.S. patents, occupations and worker earnings	Workers exposed to technologies that make their existing skills obsolete suffer displacement or weaker earnings growth. The authors explicitly emphasize vintage-specific human capital. (National Bureau of Economic Research)	Provides the theoretical mechanism: experience can become a liability when it is tied to an obsolete technological vintage. NBER study
Barth, Davis, Freeman & McElheran (2020), “Twisting the Demand Curve”	U.S. firms' software investment	Software investment raises earnings, but the effect declines after age 50 and is approximately zero after age 65; conventional equipment investment does not show the same age pattern. (National Bureau of Economic Research)	Evidence that digital technology can have an age-gradient in its wage effects. NBER study
Bartel & Sicherman (1993), “Technological Change and the Careers of Older Workers”	35 U.S. industries	Expected technological change induces more training and later retirement, but an unexpected increase in technological change causes older workers to retire earlier, apparently because retraining becomes less attractive. (National Bureau of Economic Research)	Direct historical evidence that unexpected technological disruption can cause older workers to exit rather than retrain. NBER study
Braxton & Taska (2023), “Technological Change and the Consequences of Job Loss”	U.S. occupational skills and displaced workers	Technological change explains about 45% of the decline in earnings following job loss. Workers who lack the new skills move to occupations where their remaining skills command lower wages. (American Economic Association)	Strong evidence for your second proposition: replacement employment can be less lucrative, even when workers remain employed. American Economic Review study
Love & Torrence (1989), “The Impact of Worker Age on Unemployment and Earnings After Plant Closings”	Older vs. younger displaced U.S. workers	Workers 55+ had a median unemployment duration of 27 weeks versus 13 weeks for workers under 45, and older workers subsequently earned less. (PubMed)	Older workers have historically had more difficulty converting displacement into equivalent new employment.
2001 Journal of Socio-Economics study, “Age bias in worker displacement”	U.S. displaced workers	Older displaced workers had higher pre-displacement wages but suffered greater earnings losses than younger displaced workers. (ScienceDirect)	Particularly relevant to the idea that an older worker may be unable to reproduce the economic value of a lost job.
Hudomiet & Willis (2021), “Computerization, Obsolescence, and the Length of Working Life”	U.S. computerization, 1984–2017	Older workers initially adopted computers later than younger workers, creating a temporary knowledge gap; computer use eventually converged. (National Bureau of Economic Research)	Suggests that the problem is not an inherent inability of older workers to learn technology; the timing of adaptation matters. NBER study

The point: fear of new technology impact on job prospects is an old story. What seems unusual today is just that the most-technologically-adept generations are those who now fear the implications AI has for jobs, even if they are most likely to get new jobs created by AI.