AI is not a reason to shrink.

Jensen Huang said something this week that every business owner needs to hear.

He called out companies using AI as an excuse to cut headcount. His word for it: a lack of imagination.

His point is simple. If AI makes your team faster and more capable, the move is not to shrink. It is to do more. Build more products. Serve more customers. Enter markets you could not reach before.

The radiologist that nobody replaced

The radiologist example makes it real.

For years, the narrative was clear: AI would replace radiologists. The technology could read scans. Why would you need the human?

What actually happened: demand for radiologists went up.

Because AI made each one more productive, more scans could be read per doctor, more patients could be served, more diagnostic capacity could be created. The bottleneck shifted. Instead of fewer radiologists, the system needed more throughput — and that required more people operating at a higher level.

Efficiency did not shrink the profession. It expanded what the profession could do.

The Jevons Paradox

Economists have a name for this.

In 1865, William Stanley Jevons noticed something strange about steam engine efficiency. As engineers made coal-powered engines more efficient, they expected coal consumption to drop. It did not. It went up — dramatically.

Because cheaper, more efficient energy made more uses economically viable. Factories that could not afford to run at full power before could now run longer. New industries emerged that would not have existed without the efficiency gain.

The Jevons Paradox: when you make something more efficient, total demand for it tends to rise, not fall.

It happened with coal. It happened with computing — cheaper processing did not reduce the demand for computers, it created entirely new categories of use. It is happening with radiology right now.

And it is about to happen with every team that figures out how to use AI properly.

The wrong question most businesses are asking

Most of the businesses I talk to are still stuck in the same frame. They see AI as a cost play.

The logic goes: AI can do some of what my team does. Therefore I can have fewer people. Same output, lower cost.

That math is not wrong on its face. But it is the wrong question.

The smarter question is: same people, what new output becomes possible?

• If your sales team can respond to every lead in minutes instead of hours, what does that do to close rates?
• If your ops team stops drowning in manual coordination, what projects do they finally have capacity to run?
• If your founder stops being the fallback for every decision, where does that freed attention go?
• If your customer success team can handle twice the accounts without dropping quality, what does that do to revenue per head?

None of those outcomes come from cutting. They come from expanding what the same people can do.

Where the leverage actually is

Cutting headcount feels like efficiency. It shows up cleanly on a spreadsheet.

But it leaves the constraint in place. You still have the same capability ceiling — just fewer people bumping against it. You have not changed what your business can do. You have reduced who is doing it.

The leverage is not in cutting. It is in expanding the scope of what each person can execute.

An operator who used to spend half their week on reporting now has half their week back. That is not just a time saving. That is a capability that was already on your payroll and was previously buried under low-value work.

AI done right surfaces hidden capacity. It does not eliminate it.

What Jensen Huang actually said

Huang's critique was not just about strategy. It was about imagination.

When a founder looks at AI and sees a headcount reduction opportunity first, they are revealing something about how they think about growth. They are optimizing for the same output at lower cost rather than asking what becomes possible that was not possible before.

The businesses that are going to look very different in three years are not the ones that ran the same playbook with a smaller team. They are the ones that looked at their current team, layered in genuine AI capability, and asked: now where do we go?

That is the right question. Most businesses are not asking it yet.

nVelocity point of view

The businesses gaining ground right now are not cutting their way to efficiency. They are expanding their operational surface.

They are using AI to do things that were previously too slow, too expensive, or too dependent on a single person. They are taking markets that were out of reach and making them accessible. They are building output that would have required a team twice the size twelve months ago.

AI is not a headcount argument. It is a capability argument.

The Jevons Paradox is not a prediction. It is a pattern. When something gets dramatically more efficient, demand for it expands — and so does what it enables.

Your team is the radiologist. The question is whether you are going to use AI to read more scans or to prove the machine was right to replace them.

That is a choice. And it starts with imagination.