Your Salary Is an Average. That's the Whole Problem.

Nobody tells you what a salary actually is

A salary is not a measure of how good you are. It is a band. Someone decided, before you ever walked into the room, roughly what this job is worth to this company, and you were slotted into it. That number was set by benchmarking against other companies paying other people to do roughly the same thing. Which means your salary is, quite literally, an average. It is the price of average work in that role. And that is fine, or at least it used to be, because for decades the deal was that if you got better, the average moved with you. That deal is quietly breaking.

What AI actually did to pay

The economists have a phrase for what is happening and it is more useful than most of the AI discourse. Knowledge, expertise and efficiency are becoming abundant. AI made them abundant. And abundant things get cheap, that is not a moral judgement, it is just what happens to the price of anything there is suddenly a great deal of. So average wages get tied more and more tightly to average, replaceable work, while premium pay accrues to the people creating disproportionate value. The premium did not disappear. It moved. And it moved somewhere a salary structurally cannot follow it, because a band is built to pay the average, not the exception. You can be the exception inside the band and still get paid the band.

Why raises stopped feeling like raises

Look at what people actually do with a raise now. Thirty-seven percent save it. Thirty-two percent put it straight onto basic expenses. Almost half of workers say they do not believe their wages will ever catch up with the cost of living, and only about one in five think wages will catch inflation this year. That is not pessimism, that is people correctly reading a chart. Since the late seventies productivity has climbed, profits have climbed, and median wages have gone more or less sideways once you adjust for inflation. The gains went somewhere. They went to owners and shareholders. So when you negotiate hard and win a good raise, you are not moving up, you are staying level, and you are staying level inside a structure that was never designed to pay you a premium in the first place.

The workflow: find out what your disproportionate value actually is

Here is the useful part, and it uses AI to do the thing AI is genuinely good at, which is holding up a mirror.

First, open Claude or ChatGPT and give it an honest inventory of what you actually do in a week. Not your job description. The real list, including the invisible stuff, the calls you take, the decisions you get asked to make, the fires you put out.

Second, ask it to sort that list into two piles. Work that is now abundant, meaning a competent person with the right tools could produce a passable version, and work that is genuinely scarce, meaning it depends on your judgement, your relationships, your accountability, or your read on a situation nobody briefed you on. Be brutal about the first pile. Most of us overestimate how rare our competence is.

Third, take the scarce pile and ask it the money question. For each item, what would a company pay to have this done well by someone who is not on their payroll. Not what your salary implies it is worth. What it is worth on the open market, to someone who needs exactly that and cannot get it internally. Have it give you a range and a reason.

Fourth, compare that number to what you are actually paid for it. In almost every case, the answer is that your scarce work is subsidising your abundant work. You are being paid an average for a portfolio of things, some of which are worth far less than the average and one or two of which are worth far more.

That gap is the whole game. It is the thing a salary is structurally designed to hide, because a band pays for a role, not for the two things you do that nobody else can.

What to do with the gap

Knowing the number does not oblige you to do anything dramatic. But it does change what you are optimising for. Once you can see that one slice of your work commands a premium and the rest is being repriced downwards by a technology that is not going away, spending another year fighting for four percent on the whole average starts to look like a strange use of your energy. The people who do well from here will not be the ones who won the biggest raise. They will be the ones who worked out which part of what they do is genuinely scarce, and then found a way to sell that part directly, at its actual price, rather than having it averaged into a band with everything else.