The numbers come first. Nearly four years after ChatGPT's launch in late 2022, occupations with the greatest exposure to AI have not experienced widespread employment declines, according to Vanguard Chief Economist Adam Schickling. Writing in Fortune, Schickling argues that employment growth in highly AI-exposed occupations has 'generally kept pace with — or exceeded — that of less exposed occupations.' Layoff rates remain low, and any hiring slowdown has been broad-based, not concentrated in AI-intensive fields.
Schickling reaches back to the ATM era to make his case. When automated teller machines became widespread in the 1980s, most forecasters assumed bank tellers were finished. The reality was more nuanced. ATMs lowered branch operating costs enough that banks found it economical to open more locations. Total U.S. bank teller employment remained broadly stable from 1980 through 2010. The work inside branches shifted up the skill-value chain — away from processing transactions and toward managing customer relationships — while banks added loan officers, credit analysts, personal bankers, and fraud and risk specialists.
The genuine disruption, Schickling notes, arrived later and from a different direction. Mobile banking, accelerating after 2010, did not merely automate a task — it automated the entire trip to the branch. By 2025, only 9% of bank customers said branches were their primary banking channel, down from 36% in 2007. Bank teller employment fell accordingly. Crucially, that shift required more than technology alone. The Electronic Signatures in Global and National Commerce Act of 2000 gave electronic signatures the same legal standing as ink signatures, helping enable fully digital banking and accelerating the move away from in-person transactions.
The structural lesson Schickling draws is direct: 'Isolated task automation rarely results in large-scale job losses, except in occupations built around a very narrow set of activities.' Meaningful disruption, he writes, occurs when technologies combine with new workflows, business models, and institutional changes that fundamentally alter how work is organized. Mobile banking's disruption also created entirely new employment categories — cybersecurity analysts, digital product managers, payment-platform engineers, and data-platform operators.
If AI follows the trajectory of electricity or the personal computer — general-purpose technologies that enabled products and industries not yet envisioned at the time of their introduction — the fears of white-collar job destruction are, in Schickling's words, 'likely overblown.' The more significant labor-market reconfiguration, he argues, may require something closer to the ATM-to-mobile-banking leap: a deeper redesign of business processes, organizational structures, and customer interactions. 'The evidence today suggests we remain closer to the ATM phase than the mobile banking phase.'
CEO Times take: The market has already voted on this question, and the vote aligns with Schickling's reading. Capital continues to flow into AI-adjacent businesses precisely because productivity gains — not mass displacement — are the dominant near-term story. What the ATM analogy ultimately vindicates is the free-enterprise instinct: let technology and market incentives reorganize work, and the labor force adapts upward, not downward. The danger is not the algorithm. The danger is a regulatory overreaction — new mandates, AI-use restrictions, or preemptive labor rules — that freezes the ATM phase in place and prevents the productivity dividend from ever arriving.
