In 1865, the English economist William Stanley Jevons noticed something that did not add up. Coal-fired furnaces were getting more efficient. Logic said that should mean less coal burned. Instead, consumption climbed. Cheaper, more efficient coal use made coal more attractive, demand spiked, and overall spending on coal rose rather than fell.

That observation became known as the “Jevons Paradox.”

It is a useful lens for the question hanging over financial service jobs right now: what happens to the jobs when AI gets good and cheap?

The Predicted AI Jobpocalypse

In early 2025, Goldman Sachs predicted that AI could displace 6 to 7% of the US workforce under wide adoption. Around the same time, the World Economic Forum projected that AI could displace 92 million jobs. However, that same World Economic Forum report had another interesting number: AI could simultaneously create 170 million new jobs.

In life, the only inevitable is change.

Will AI cause some change for wealth managers, accountants, and CFOs? Probably. The Internet changed things. Microsoft Excel changed things. And AI will likely change things too.

However, it is worth examining the leap from “the work will change” to “the jobs will disappear.” Those are not the same claim, and history is unkind to the second one.

It Happened Before

The AI-driven job destruction argument is not new. Back in 2016, Geoffrey Hinton, one of the godfathers of modern AI, predicted that AI would make the radiologist profession obsolete within five to ten years. It was a confident call from someone with every credential to make it.

Fast forward to 2026. The displacement never came. The U.S. Bureau of Labor Statistics now projects “faster than average” job growth for the field from 2024 to 2034. AI did not replace radiologists. It let them do more.

Enable an expert to work more efficiently and the demand for the work tends to grow, not shrink.

An Additional Wrinkle

There is a second force at play in financial services. The supply of experienced professionals is thinning out.

In the field of wealth management, Cerulli Associates found that over the next ten years, 37.4% of wealth managers plan to retire. Similar pressure is building in accounting and across the ranks of seasoned financial professionals.

So the real near-term problem in not that firms may too many people for the work. It is that they may have too few. In that setting, AI is even less of a threat to jobs and more like a way to keep up with demand as experienced hands walk out the door.

‘Invision’ the Possibilities

There is a third trend unfolding, and it may be the most interesting one. AI does not just meet existing demand or backfill retirements. It opens up work nobody had thought of.

A 2025 LinkedIn study found that 20% of people in the US hold job titles that did not exist in the year 2000. Twenty years ago, those roles were unimaginable. Today they are someone’s career.

What does the next 1, 5, and 10 years look like for financial services? It might be a future with less tedious work and more time on the work that actually adds value. It might be a future where interpersonal skills matter more, not less. And it might be a future where AI propels people to new heights rather than replaces them.

If the Jevons Paradox holds, as it has for a century and a half, then AI could ultimately be a nice tailwind. Not a headwind.