AI and Job Losses: Are We Looking at This All Wrong?
You're Asking the Wrong Question About AI and Jobs
The conversation about AI and employment keeps getting stuck in the same place.
One side predicts catastrophe. The other cites historical precedent and tells everyone to relax. Both are partially right, which means both are missing something important.
The question is not "Will AI take jobs?" It will. The real question is: Who gets hit first, how fast does it happen, and are you positioned on the right side of that line?
The Historical Argument Is Real -- and Incomplete
The optimists have a case worth taking seriously.
MIT economist David Autor and colleagues published research in the Quarterly Journal of Economics (2024) showing that roughly 60% of today's workers are employed in occupations that did not exist in 1940. More than 85% of net job growth over the past eight decades came from entirely new categories of work. Marc Andreessen, writing in his 2023 essay "Why AI Will Save the World," frames the optimist view directly: "AI, if allowed to develop and proliferate throughout the economy, may cause the most dramatic and sustained economic boom of all time, with correspondingly record job and wage growth." -- Marc Andreessen, a16z.com (2023)
That argument is grounded in real economic logic. When productivity rises, prices fall, demand grows, new industries form. It has happened before.
But here is what that argument cannot tell you: who absorbs the cost during the transition, and how long that transition actually takes. The Industrial Revolution took roughly 60 years before productivity gains reached average workers. The internet gutted print media and retail over 20 years. Those were painful timelines measured in careers, not economic cycles. Most people citing historical precedent are comforting themselves with the long run. The short run is where real people live.
The Warning Coming From Inside the Industry
When the CEO of an AI company warns that his own technology could cause a serious employment crisis, that statement is worth examining carefully.
Dario Amodei, CEO of Anthropic, told Axios in May 2025 that AI could eliminate up to 50% of entry-level white-collar jobs within five years -- with unemployment potentially rising to 10--20%. He was specific: finance, consulting, law, and tech. Entry-level, structured, digital work. Speed: "almost overnight" once business leaders see the savings of replacing humans with AI.
"We, as the producers of this technology, have a duty and an obligation to be honest about what is coming. I don't think this is on people's radar." -- Dario Amodei, CEO of Anthropic, Axios (May 2025)
What happened in the year after that warning? Entry-level technology hiring at major firms dropped an estimated 30-50% compared to 2023. Nearly 55,000 U.S. job losses in 2025 were directly attributed to AI, by the companies cutting them. Workday, Amazon, Meta, Google, Salesforce, Microsoft, Shopify. The language was consistent: "AI enables leaner structures."
This is not a prediction anymore. It is a pattern.
The Gap Nobody Wants to Name
AI will, over a long enough horizon, likely create more economic value than it destroys. Goldman Sachs Research found that while AI could displace 6-7% of the U.S. workforce under baseline assumptions, historical patterns suggest demand for new work eventually absorbs displaced workers. The macro-level optimism is probably correct.
The tradeoff is: that is a 10-to-20-year story. The disruption hitting entry-level white-collar workers is a 2-to-5-year story. Those timelines do not overlap in a useful way for the people caught in between.
Here is what that looks like in practice. A 22-year-old finishing a finance degree right now is entering a market where the tasks that used to constitute the first two years of their career (i.e. document review, financial modeling, earnings summaries, due diligence research) are being done by AI agents at a fraction of the cost. The new jobs AI creates exist. But they require skills that take time to develop. That gap is not a minor inconvenience. It is a structural hole in the middle of the career ladder.
What This Actually Demands
If you are running a team or an organization, the honest question is: which roles in your structure are doing work that AI can already do reliably at lower cost? The answer is almost certainly broader than you want it to be. The companies that figure this out early will redesign their talent structures deliberately. The ones that wait will do it reactively, under financial pressure, in a way that is worse for everyone involved.
If you are an individual contributor doing routine, structured, digital work -- legal review, financial analysis, code generation, content production, customer support -- the pattern-recognition window is open right now. It will close. The people who move before the wave are in a different position than the people who move because of it.
The historical record says new work will be created. It does not say it will be created on a schedule that is convenient for your specific situation.
History is on the side of long-run optimism. The short run is still being written. The difference between those two timelines is where most people are currently standing without realizing it.
About the Author
Martin Goetzinger has spent his career in enterprise software sales, helping large organizations such as Apple, Microsoft, and Verizon connect data, insight, and action. His work focuses on transforming how businesses measure success and create customer value through technology.
Outside the enterprise world, he writes about the five forces he believes are reshaping everything: AI, blockchain, energy, personalized health, and robotics. Not from a purely technical lens, but from a human one as to how these technologies will redefine work, wealth, and well-being.
He is based in the U.S. and publishes at www.MartinGoetzinger.com.
