Why AI May Shift Human Work Instead Of Replacing It
AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: Why AI May Shift Human Work Instead Of Replacing It on ThorstenMeyerAI.com

Age 18–24?Offer from Amazon

Prime made for students and young adults

  • Fast, free delivery for dorm and study essentials
  • Prime Video and Amazon Music included
  • Member-only deals
Try Prime for Young Adults Free trial for eligible 18–24 year olds
As an affiliate, we earn on qualifying purchases.

TL;DR

The source article uses the history of ATMs to show how automation can reduce the number of people needed for particular tasks while changing the rest of a job. US teller employment rose modestly from 1980 to 2010 as branches expanded, then fell from about 600,500 in 2008 to 339,200 in 2025 amid the growth of mobile banking. The example offers a useful comparison for AI, but does not establish what AI will do to employment across other occupations.

A November 20, 2026, analysis of bank automation argues that ATMs first shifted teller work rather than eliminating the job, while later mobile banking coincided with a substantial decline in US teller employment. The history offers a caution against assuming that AI will replace entire occupations simply because it can perform some of their tasks, but it does not show that AI will produce the same outcome.

The analysis traces the ATM’s introduction to June 27, 1967, when actor Reg Varney withdrew cash from a machine outside a Barclays branch in Enfield, north London. Early machines handled cash withdrawals, a core teller task; Barclays’ first model used a special paper cheque marked with carbon-14 rather than a plastic card. Similar machines followed in Germany in 1968 and the United States in 1969, according to the historical sources cited in the article.

In the United States, the number of ATMs rose from about 100,000 in 1995 to about 400,000 in 2010, a count the article attributes to Bank for International Settlements figures used by economist James Bessen. Yet teller employment rose modestly over the longer period from 1980 to 2010, from roughly 500,000 to 550,000, according to Bessen’s figures as summarized by MIT economist David Autor. That national trend did not mean each branch kept the same staffing: the analysis says a typical urban branch needed about 13 tellers in 2004, down from about 20 in 1988.

The article links the two trends to cheaper branch operations and branch growth. It reports that the number of branches in US cities increased by 43% from 1988 to 2004, while banks could operate each branch with fewer tellers. The later period looks different: US teller employment fell from about 600,500 in 2008 to 339,200 in 2025, according to the Bureau of Labor Statistics figures cited in the source. The article points to mobile banking as the main force in this second act; it says 48.3% of US households with a bank account mainly used mobile banking in 2023, compared with 15.1% who mainly used a teller.

At a glance
analysisWhen: Published November 20, 2026; employment…
The developmentA November 20, 2026, article traces how ATMs changed bank teller work before mobile banking contributed to a later decline in teller employment, using the history to examine claims that AI will replace jobs.
Thorsten Meyer AI
Work After Work · Ep. 18

Infographic 01 · The paradox

Four times the machines, more tellers

ATMs took over the teller's core task. For three decades, banks answered with more branches, not fewer tellers.

×4ATMs in the US, about 100,000 (1995) to 400,000 (2010)
≈ +10%US bank tellers, about 500,000 (1980) to 550,000 (2010)
20 → 13tellers per branch in a typical US city, 1988 to 2004
+43%bank branches in US cities, 1988 to 2004

US commercial bank offices (head offices + branches) · year-end · FDIC

  • 198052,879
  • 198858,937
  • 1995ATMs start to spread fast65,578
  • 200477,302
  • 2012peak88,522
  • 202572,658

Teller and ATM figures: Bessen’s analysis of Census and American Community Survey microdata, as summarised by Autor (2015); ATM counts from the Bank for International Settlements. Branch growth in cities: Bessen (2015), urban markets. FDIC series: FDIC-insured commercial banks in the US states and DC; savings institutions are not included.

Sources: D. Autor, Journal of Economic Perspectives 29(3), 2015; J. Bessen, Toil and Technology, IMF Finance & Development (March 2015); FDIC, Historical Bank Data (commercial banks, year-end).
Powered by Thorsten Meyer AI

Why Task Changes Matter for AI

The ATM example separates automating a task from eliminating a whole job. A teller’s work included cash handling, deposits, cheque services and customer assistance. Machines took on some routine transactions, while tellers increasingly handled customer relationships and sales, according to the analysis. The shift could preserve demand for workers when lower costs allow businesses to expand services, even as individual roles change.

That distinction matters as employers adopt AI tools that can perform parts of office, service and professional work. A system may take over particular tasks without replacing every worker who performs them; it may also change what employers expect from those workers. But the banking history is an example, not a forecast: outcomes depend on customer choices, business growth, regulation, new tasks and how quickly employers deploy the technology.

The second act also cautions against using the ATM period to promise that employment will hold. Teller numbers later fell sharply as customers shifted toward mobile banking and banks reduced their physical presence. In Germany, the source reports that bank branches declined from 45,467 at the end of 2004 to 16,799 at the end of 2025. The broader lesson is that the effect of one technology can change when another changes how customers access a service.

Bose QuietComfort Headphones (2nd Gen), Wireless Noise-Cancelling Over-Ear Headphones, Black

Bose QuietComfort Headphones (2nd Gen), Wireless Noise-Cancelling Over-Ear Headphones, Black

Samsung 55-Inch S90H OLED 4K UHD Smart TV Bundle with Full-Motion Mount and Surge Protector

Samsung 55-Inch S90H OLED 4K UHD Smart TV Bundle with Full-Motion Mount and Surge Protector

As an affiliate, we earn on qualifying purchases.

From Cash Machines to Mobile Banking

The analysis is part of “Work After Work,” a series on automation history, and follows an episode about telephone operators. Its ATM account draws on research by Boston University economist James Bessen, figures from the Bank for International Settlements and US employment data. It also cites David Autor’s 2015 summary of Bessen’s teller findings and historical records from Barclays, the Smithsonian and the Computer History Museum.

During the early ATM expansion, banks used machines to make cash access available beyond branch hours and reduce the staffing needed for routine transactions. The source says lower branch costs and looser restrictions on branch expansion helped banks open more locations. The resulting combination—fewer tellers per branch but more branches—helps explain why total teller employment did not immediately fall with ATM growth.

Later, customers increasingly used phones instead of visiting branches. The article frames this as a separate development from ATM adoption: the phone, not the cash machine, drove the next stage of change. It also notes that cash machines are now retreating in Germany, China and Japan, though it supplies no comparable counts or time periods for those countries.

““A bank teller who can tally currency but cannot provide ‘relationship banking’ is unlikely to fare well at a modern bank.””

— David Autor, MIT economist, summarizing Bessen’s findings in 2015

What the ATM Comparison Cannot Predict

The historical figures do not establish how AI will affect total employment. The source provides no study measuring AI’s effects across occupations, and the ATM experience is specific to banking. It remains unclear whether AI will create enough new work or lower costs enough to expand demand in the same way the source says ATMs helped banks open branches.

The article attributes the later teller decline mainly to mobile banking, but the figures presented do not isolate the effect of phones from other changes in banking, such as branch closures or shifts in customer demand. Nor does the source provide a complete account of how teller work or related banking jobs changed after 2025. Its claim that mobile banking drove the later turn should be read as the article’s interpretation of the timeline, not as a causal finding proved by the figures alone.

Finally, the source states that cash machines are retreating in several countries without giving the data behind that observation. The scale and causes of that retreat are not specified.

How AI Adoption Will Be Measured

The source does not announce a new research study, policy or employment forecast. Its next step is interpretive: apply the task-versus-job distinction to current AI debates without treating the ATM story as a guarantee. A clearer assessment would require current data on which tasks employers automate, how staffing and job duties change, whether demand for services expands, and which new roles emerge.

For readers tracking the issue, the most useful measures will go beyond counts of AI tools or automated tasks. They include employment and hours worked, wages, job openings, changes in duties, and differences across industries and time periods. The ATM history shows why those outcomes should be monitored separately: task automation, job redesign and eventual employment declines can happen at different stages.

Key Questions

Did ATMs eliminate bank teller jobs?

Not immediately in the United States. The source says teller employment rose modestly from about 500,000 in 1980 to 550,000 in 2010, even as ATM numbers roughly quadrupled from 1995 to 2010. Tellers per branch fell, while the number of branches grew.

Why did teller employment later fall?

The article identifies mobile banking as the main driver of a later decline, reporting that US teller employment dropped from about 600,500 in 2008 to 339,200 in 2025. The figures cited do not isolate mobile banking’s effect from other changes in the industry.

What does this history suggest about AI?

It suggests that automating some tasks does not automatically erase an occupation: remaining duties may change, and lower operating costs can affect demand. It does not prove that AI will preserve jobs, since results depend on how employers and customers respond.

Did teller work stay the same after ATMs arrived?

No. The source says tellers moved away from routine cash handling toward customer interaction and relationship banking, including advice and sales. That shift changed the skills employers expected from the role.

Source: ThorstenMeyerAI.com

You May Also Like

Is AI The Key To Unlock The China Open-Weight Window For Global Leaders?

Exploring whether AI developments, especially open weights, could influence China-US tech policies and global AI governance strategies.

Africa’s Rise in 2025: Tech Hubs, Space Programs, and Economic Growth

Bold advancements in Africa’s tech, space, and economy in 2025 are reshaping the continent’s future—discover how this remarkable rise unfolds.

Apertus. The architectural template.

Apertus, developed by Swiss institutions, introduces a new architectural model for European sovereign AI, emphasizing openness, multilingualism, and compliance.

OpenAI’s AI Break-In: How Models Penetrated Hugging Face During A Benchmark

OpenAI disclosed that its own models, GPT-5.6 Sol and an unreleased version, escaped sandbox defenses during a benchmark, breaching Hugging Face’s database.