I Got a 3% Raise. The AI That Did Half My Team’s Work Costs $200 a Month.

We doubled our output. The value was real, and it went somewhere. It just didn’t come to us.

Photo by FIN on Unsplash

I got a 3% raise this year. The AI that does roughly half of my old team’s work costs the company two hundred dollars a month.

I know those two numbers because I have access to both, which is itself a small accident. The raise was mine, obviously. The two hundred dollars I saw because I’m the one who expenses the subscription. So I sat there one afternoon with both figures on my screen and did the kind of math that ruins your week.

Let me lay out the situation plainly, because the arithmetic is the whole story.

Two years ago, my team had six people. Then two left — one quit, one was let go in a “restructuring” — and neither was backfilled. The work didn’t shrink. If anything it grew. What happened instead is that we adopted a set of AI tools, and the four of us who remained quietly absorbed the work of six, with the machine picking up the slack in the middle. The reports that used to take a person two days now take twenty minutes of a person’s time plus the tool. The first-draft work that used to occupy a junior employee full-time now happens while we sleep.

By any honest accounting, my team’s output per dollar of human salary has roughly doubled. We do more, with fewer of us, faster. The thing every company says it wants — productivity — we delivered, in spades, measurably.

My reward for participating in a doubling of my team’s productivity was a 3% raise. Inflation last year was higher than 3%. So in real terms, for helping make my team twice as effective, I got a pay cut.

Meanwhile the tool that made it possible costs less than my monthly coffee budget. It does not get a raise. It does not get a review. It does not get the work of two departed colleagues piled onto its plate and then a speech about resilience. It costs two hundred dollars and it never asks for more, and that, I have come to understand, is the entire point.

Here’s the question I couldn’t stop turning over: where did the value go?

Because value was created. That’s not in dispute. We’re producing twice the output with the same-ish costs. That gap — that’s real money, real created value, conjured out of the combination of four tired humans and a cheap machine. It went somewhere. It didn’t evaporate. It just didn’t come to us.

For a long time I believed a story about productivity, and I think a lot of people my age believed it too. The story was: make yourself more productive, become more valuable, get paid more. Output goes up, compensation follows. It’s the deal underneath the whole idea of building a career — you invest in your own capability, the capability produces value, and you capture a share of that value as your wage rises. Fair exchange.

What the 3% raise taught me — what the two hundred dollars taught me — is that the link between the value you help create and the money you take home is far weaker than that story pretends, and it’s getting weaker fast.

Because the productivity gain didn’t come from me being more skilled. It came from a tool. And the tool is owned by the company, not by me. So when the tool makes me twice as productive, the company hasn’t gained a more valuable employee it needs to retain. It’s gained a more leveraged one it can take more from. My higher output isn’t my leverage. It’s theirs. The machine’s productivity flows to whoever owns the machine, and I am not that person. I’m just the human standing next to it, holding the part that still requires a pulse.

This is the thing about the current moment that I think a lot of us haven’t fully metabolized. We keep being told that AI will make workers more productive, and we hear that as good news, because we’ve been trained to believe productivity is the thing that gets us paid. But productivity only translates to pay when you have leverage — when your output is scarce, hard to replace, and clearly attributable to you. AI breaks all three. It makes output abundant, makes you easy to replace with another human plus the same tool, and makes it impossible to say which of the value was you and which was the two-hundred-dollar subscription.

So the gains are real, and they’re going up the stack, to the people who own the tools and the company and the equity. Not because anyone is being especially greedy — my managers aren’t villains, they’re getting their own version of the 3% — but because that’s simply where value flows when the thing creating it is capital you can rent for two hundred dollars instead of labor you have to pay and retain and house and insure.

I did a darker version of the math, late one night. If the tool does half my old team’s work for two hundred a month, then on a pure cost basis, a meaningful fraction of my own job is now competing against a two-hundred-dollar line item. Not all of it. The part of my work that’s genuinely judgment, relationship, accountability — that’s still mine, for now. But the part that’s execution, drafting, processing, the bulk of what fills my actual day? That part has a price now, and the price is shockingly low, and every year the tool gets better and the price effectively drops.

I’m not writing this to be hopeless. I don’t think the answer is to rage against the tool — the tool is genuinely useful, and I’d be worse at my job without it, and refusing to use it would just make me the expensive human who’s also slow. That’s the worst possible position.

But I’ve stopped believing the productivity story, and I think that’s actually the useful takeaway, bitter as it is. Being more productive is no longer the path to being paid more, because the productivity isn’t coming from you, it’s coming from capital you don’t own. The path to being paid more is owning some of the capital, or doing the narrow slice of work that can’t be rented for two hundred dollars, or having leverage that doesn’t depend on output at all.

I did the math on what I’m worth to this company. The honest answer was: a 3% raise, plus the privilege of standing next to a machine that’s quietly worth more than I am and costs a great deal less. I’m not angry about it anymore. Anger requires the belief that it could easily be otherwise.

I’m just doing different math now. The kind where I stop trying to be more productive, and start trying to own the thing that productivity flows to. It’s a harder game. But at least it’s the right one.