How To Think About AI And The Future Of Work
AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: How To Think About AI And The Future Of Work on ThorstenMeyerAI.com

Before you orderOffer from Amazon

Get the latest gadgets delivered free with Prime

  • Fast, free delivery on millions of items
  • Prime Video, Amazon Music and more included
  • Member-only deals all year
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

TL;DR

A November 20, 2026 article from ThorstenMeyerAI.com revisits the ATM’s effect on bank tellers: teller employment in the United States rose modestly from 1980 to 2010 even as ATMs multiplied, in part as cheaper branches spread. The article argues that mobile banking and branch closures later drove a different employment decline, while cautioning that the history offers context—not a direct forecast—for AI’s effects on work.

ThorstenMeyerAI.com published an account of ATM automation on November 20, 2026, arguing that the machine did not initially eliminate US bank teller jobs: employment rose modestly from about 500,000 in 1980 to 550,000 in 2010, even as ATMs became far more common. The article says the later spread of mobile banking and branch closures brought a much steeper decline, offering a historical case study for debates about AI and work rather than a guaranteed forecast.

The article opens with the first Barclays cash machine, demonstrated by actor Reg Varney outside a branch in Enfield, north London, on June 27, 1967. The machine dispensed cash after customers inserted a special paper cheque and entered a code. It could operate after the branch closed, but it handled only part of a teller’s work: cash withdrawals and, in later machines, other routine transactions. The article also records early machines in Germany in 1968 and the United States in 1969.

Drawing on economist James Bessen’s analysis, the account says the US ATM count rose from about 100,000 in 1995 to about 400,000 in 2010. Yet teller employment increased modestly over the longer 1980–2010 period. One explanation was that machines reduced staffing needs at individual branches while lowering the cost of operating branches. In US cities, the article reports, branch numbers rose 43% from 1988 to 2004, while the typical urban branch needed about 13 tellers rather than 20.

The subsequent employment decline is linked to changing customer habits and a shrinking branch network. The article cites data showing that in 2023, 48.3% of US households with a bank account mainly used mobile banking, compared with 15.1% that mainly used a teller. It reports that US teller employment fell from about 600,500 in 2008 to 339,200 in 2025. Germany’s bank branches fell from 45,467 at the end of 2004 to 16,799 at the end of 2025, according to the source article.

At a glance
reportWhen: Published November 20, 2026; the articl…
The developmentThorstenMeyerAI.com published an episode tracing how ATMs changed bank teller work and how mobile banking later contributed to a sharper decline in teller employment.
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

How Automation Reshaped Teller Work

The history complicates a simple prediction that automating a task automatically removes the entire occupation. ATMs took over routine cash transactions, but the article’s account says banks used tellers for more customer-facing work, including advice and product sales. A technology can change the mix of tasks in a job, reduce staffing at each location, and also make a service cheaper to provide. In this case, branch expansion helped offset the need for fewer tellers per branch for a time.

That distinction matters for current discussions of AI and employment. The ATM example suggests that effects can depend on whether automation reduces costs, changes demand, or shifts workers toward tasks that remain human-led. It does not establish that AI will follow the same path: the technologies, occupations, and economic conditions differ. The later teller decline also shows that job effects can unfold in stages, as new services and customer habits alter demand.

LG OLED evo C4 55-Inch 4K Smart OLED TV (OLED55C4PUA) with 144Hz Refresh Rate, HDMI Cable, 24-Month Protection Plan, and Streaming Bundle

LG OLED evo C4 55-Inch 4K Smart OLED TV (OLED55C4PUA) with 144Hz Refresh Rate, HDMI Cable, 24-Month Protection Plan, and Streaming Bundle

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

Samsung 65-Inch S90H OLED 4K Smart TV Bundle with HW-B630F Soundbar, Subwoofer, and Screen Cleaner (2026)

Samsung 65-Inch S90H OLED 4K Smart TV Bundle with HW-B630F Soundbar, Subwoofer, and Screen Cleaner (2026)

As an affiliate, we earn on qualifying purchases.

From Cash Machines to Mobile Banking

The first ATM arrived when customers generally had to visit a bank during office hours to get cash. The early machines were designed around a narrow, visible task, and their name—automated teller machine—reflected the work they were meant to handle. As cash transactions moved to machines, teller roles shifted toward helping customers and building banking relationships.

The article separates that first phase from a later one driven by phones and online services. It cites Bessen’s work on the period when ATMs expanded and branches multiplied, then points to mobile banking and declining branch numbers as part of the later change. In the United States, the FDIC’s branch-office count rose from 58,937 at the end of 1988 to 77,302 at the end of 2004, before reaching a reported peak of 88,522 in 2012. These figures describe different measures and periods; they should not be read as a single, uninterrupted employment trend.

“Actually rose modestly from 500,000 to approximately 550,000 over the 30-year period from 1980 to 2010.”

— David Autor, summarizing James Bessen’s findings in 2015

What the ATM Analogy Cannot Predict

The article uses bank tellers as a historical comparison, not as evidence that AI will create the same employment pattern. It does not quantify how AI may affect particular occupations, how quickly employers will adopt it, or whether new demand will offset tasks and roles that are reduced. The source material also gives broad employment figures and selected dates, but does not provide a full breakdown of teller work by task, region, or employer.

The reasons behind the later teller decline cannot be assigned to mobile banking alone from the figures provided. Branch closures, changing banking services, and other industry changes may also have played a role. The article notes that cash machines are now retreating in Germany, China, and Japan, but supplies no counts or time windows for those changes. Their scale and causes remain unclear from the source.

Track Tasks, Demand and Jobs

The article’s practical test for future automation claims is to look beyond the number of machines or software tools. Readers and policymakers will need to track which tasks are automated, whether customers use more or less of the service as costs change, and whether workers move into other responsibilities. For banking, further evidence would include updated teller employment figures, branch counts, and data on how customers divide their activity among mobile, online, ATM, and in-person services.

For AI, the same measures will need to be gathered by occupation and over time. The source article does not announce a new employment forecast or policy decision. Its next step is a historical comparison: the teller example shows why job outcomes depend on more than a technology’s capacity to perform one task, while leaving AI’s eventual effects unresolved.

Key Questions

Did ATMs immediately eliminate bank teller jobs?

No. The article cites US teller employment rising modestly from about 500,000 in 1980 to 550,000 in 2010, a period when ATM numbers also grew substantially. The figures do not mean every bank or community saw the same pattern.

Why did teller employment hold up as ATMs spread?

According to the account, ATMs reduced the number of tellers needed at each branch, but they also made branches cheaper to operate. US city branch numbers grew 43% between 1988 and 2004, helping offset lower staffing per branch.

What drove the later decline in teller jobs?

The article points to mobile banking and a decline in bank branches as part of the later shift. It reports US teller employment falling from about 600,500 in 2008 to 339,200 in 2025, but the supplied figures do not isolate the effect of each cause.

Does the ATM story show what AI will do to jobs?

No. It shows that automating particular tasks does not by itself determine what happens to an entire occupation. The effects of AI remain uncertain and will depend on adoption, changes in demand, and how work is reorganized.

Source: ThorstenMeyerAI.com

You May Also Like

Maersk Says Legacy System Was Involved In Data Theft

Maersk says attackers accessed and extracted data from a local system inherited through an acquisition. The investigation continues; a ransomware group’s claims are unverified.

How AI Powerhouses Like Granite 4.2 LLMs Are Crafted From The Ground Up

IBM unveils Granite 4.2, a family of dense reasoning language models in 3B, 8B, and 30B sizes, supporting tool calls and reinforcement learning.

Understanding The AI Apocalypse Through Grok And Claude’s Eyes

A Nautilus article prompts Grok and Claude to discuss AI risks, raising questions about chatbot responses and their implications for AI safety debates.

Exploring AI-Driven Real-Time Intelligence With IBM Time Series Models On Confluent

IBM Granite Time Series models now available in Early Access on Confluent Cloud, enabling real-time forecasting and anomaly detection within Apache Flink.