A few months ago I wrote about AI as a force multiplier and what that actually means for the job market. The customer support team that goes from fifty people to five. The engineering org that ships faster with half the headcount. I talked about how upskilling alone won't save us because the math doesn't work. More productive people means fewer people needed for the same output.

That argument still holds. There are plenty of layoffs still showing up in the news to prove it. But something has been bothering me about the way it's playing out. And I think it reveals a deeper problem with how the market is thinking about AI.

Here's what should have happened. AI dramatically increases the output a single person or a small team can produce. Products that couldn't justify a team of ten can now be built by two. Research that would take a department six months can be done in weeks. Markets that were too small to pursue become worth exploring. The logical response to a massive increase in capability per person is to pursue more things, not fewer.

But that's not what's happening. Most companies are using AI to reduce headcount while maintaining roughly the same output. They're cutting teams, pocketing the savings, and calling it a win. The quarterly earnings look great. Wall Street rewards it. And nobody seems to be asking the obvious follow-up question.

If your remaining team can now produce what your full team used to produce, what are they doing with the leftover capacity?

The answer, overwhelmingly, is nothing. They're doing the same work with fewer people. That's it.

Maintaining the Status Quo Is Not a Strategy

Companies spent years building organizations around the idea that certain things required certain levels of investment. Twenty engineers to maintain this product. A marketing team of fifteen to run these campaigns. A support org of fifty to handle this volume. AI comes along and suddenly you can do all of that with a fraction of the people.

The lazy response is to fire people and keep doing exactly what you were doing before. And right now, that's the dominant response. The market is actively rewarding companies for cutting costs rather than asking what they could be building with all this newly available capacity.

When you optimize purely for efficiency, you end up with a leaner version of the same company doing the same things. You haven't taken advantage of anything. You've just made the existing business cheaper to run. Meanwhile, all those adjacent markets that didn't quite justify the investment, all those product ideas that got killed in prioritization, all that stuff is still sitting there, waiting. Except now the economics to go after them actually work.

What Expansion Should Look Like

If a company cut its engineering team from 200 to 100 and shipped the same roadmap, I'd argue they made the wrong move. The interesting version is keeping 200 and shipping twice the roadmap. Or keeping 100 and giving them the mandate to go after the opportunities that were never on the table before.

If a team of five can now do what a team of twenty used to do, the right question isn't "how do we save the cost of fifteen people." It's "what can we now do with four teams of five that we never could have done before."

One approach gets you a cheaper version of last year. The other gets you into markets and products your competitors haven't even considered yet.

Why This Matters Beyond Corporate Strategy

This isn't just a business observation. It connects directly to the job market problem I wrote about last time.

If companies used AI to expand into new products and new markets, the people displaced by efficiency gains in one area could be redeployed into others. New products need people to sell them. New markets need people to support them. Growth creates jobs that efficiency destroys.

But when every company is optimizing for cost reduction, you get the worst of both worlds. Jobs disappear and no new jobs appear because nobody is using the gains to do anything new. The AI productivity dividend flows entirely to the bottom line instead of being reinvested.

That's the scenario I worry about. Not because AI is inherently bad for jobs. But because the way the market is responding to AI right now is bad for jobs. The technology itself is neutral. But people don't see it that way. The public sentiment around AI is largely negative right now, and honestly it's hard to blame anyone for feeling that way when the most visible outcome is people losing their jobs. AI gets treated as the villain in that story, even though the real problem is how companies are choosing to use it. Every round of layoffs attributed to "AI-driven efficiency" makes the narrative worse and makes people more resistant to the technology altogether. That's a shame, because the technology isn't the problem. The decisions are.

The Window Is Open

I keep thinking about the last time something like this happened. The early internet era had a similar dynamic. The first instinct for most businesses was to take their existing operations and put them online. Digitize the catalog. Put the brochure on a website. Use email instead of fax. The companies that did only that saved some money and moved a little faster, but they didn't build anything new. The ones that actually changed the game were the ones who looked at the internet and asked "what can we do now that was never possible before?" And those companies built things that the catalog-digitizers never saw coming.

We're at that same inflection point with AI. And most of the market is in catalog-digitizer mode. Doing the same things, slightly cheaper, with fewer people. It works for a quarter or two. It makes the board happy. But it's leaving the field wide open for anyone willing to actually use this moment to build something new.

History is pretty clear on what happens next. The companies that merely optimized eventually got overtaken by the ones that expanded. I don't think this time will be any different. The only question is how much talent and potential gets wasted in the meantime while the market figures that out.