AI Absorbs the Work Below It. Scale Absorbs the Work Above It. The Middle Is Where You Get Paid.

The conversation about AI and work has been flattened into a single bad question. Will AI replace humans, or will humans use AI as a tool. The answer to both halves is yes. Both halves miss the part that matters.There is a specific layer in every function and every career that is being squeezed from both sides at once. AI is absorbing the work below it. Scale, the increasing dominance of a few large companies in every category, is absorbing the work above it. The layer in between is where authority and compensation compound. It is the layer most companies are mispricing in both directions. And it is the layer your career, your function, and your business should be designed to occupy.I am calling this the Judgment Layer. The rest of this piece is about why naming it changes how you allocate the next two years.

 

The layer below: routine work being absorbed by AI

The work being absorbed by AI is routine cognitive work that does not require irreversible judgment under stakes. Drafting standard documents. Summarizing meetings. Generating first-draft analyses. Pulling structured information from unstructured sources. Customer service for predictable inquiries. HR operations. Junior project management. Routine code generation.

This is the work where the cost of a wrong output is low, the variance across humans doing it is low, and the upper bound on quality is reached quickly. It is being absorbed by AI faster than the discourse has caught up with. The honest companies doing it (Salesforce with customer service, IBM with HR operations) are not pretending. The phantom-AI-layoff companies absorbing this work and calling it something else are creating the cover story for it.

The layer above: work absorbed by scale

The work above the Judgment Layer is the work absorbed by scale instead of by skill. Distribution. Platform effects. Capital allocation at the largest scale. Brand recognition at the largest scale. Once a small number of companies in any category have crossed a certain scale threshold, the work above the Judgment Layer increasingly accrues to them not because they are smarter but because they are larger.

This is what is squeezing the mid-market in every industry, and it is happening faster than most operators realize.

 

The middle: judgment under stakes

The Judgment Layer is the work in between. Judgment under stakes, where the cost of a wrong call is asymmetric, where the variance across humans doing it is large, where institutional credibility is the thing that produces the outcome.

The board judgment that the equity story has to change. The clinical judgment that an AI flag is a false positive. The strategic judgment that the company should pass on the deal. The hiring judgment about the executive who looks good on paper. The marketing judgment about which positioning will compound and which will collapse. The investor judgment about which founder to back.

Here is the moment I think about when I write about this layer. The version that does not breach anyone. A senior leader I work alongside walked into a board meeting last year with a recommendation the data did not yet support. The model said hold. The market said sell. The CEO said buy. She said something more specific: buy this asset, in this configuration, at this price, only if these two things hold by the next board meeting. She was wrong on one of the two checks and right on the deal. Six months later the board reviewed the call and the consensus was that the deal worked because she had named the conditions under which it could fail, not because the model said it would succeed.

That is the Judgment Layer. The work AI cannot do because the call requires synthesis across what the data shows, what the room is reading, what the founder has not said, and what the regulator will tolerate.

The middle is where you get paid.

Why the compensation compound is the part most people miss

The pay scale is widening fastest in the Judgment Layer because the layer below is being commoditized by AI and the layer above is being commoditized by scale. The person who can hold the Judgment Layer call (with the credibility, the synthesis, the institutional knowledge, the regulatory awareness, and the cross-domain literacy that the call requires) is being paid more, and will continue to be paid more, relative to every other layer of work.

The people being squeezed out are the people in the layer below who do not move up to Judgment Layer work, and the layer above whose work is absorbed by platform dynamics they cannot fight.

 

What this changes about the next two years

If you are an operator, the question is whether the work you spend the bulk of your time on is in the Judgment Layer or in the layer below it. If it is below, the work to do is to climb into the Judgment Layer faster than the layer below you collapses. Not in three years. This year.

If you are a CEO, the question is whether the talent strategy is set up to identify and develop Judgment Layer operators across the company, or whether the company is still hiring and promoting against the layer-below criteria that worked in the previous decade. Most companies are still doing the second.

If you are an investor, the question is whether the leadership benches in your portfolio companies have Judgment Layer depth at the right places, or whether the company has one founder doing all the Judgment Layer work and a team underneath that is mostly absorbing the layer below. Most portfolio companies are doing the second, and the founder is the single point of failure as a result.

 

The counterargument worth conceding

The Judgment Layer is harder to teach than the layer below. The honest version of this argument is that we do not yet know how to develop Judgment Layer operators at scale, which means the bottleneck on the new economy is not AI capability or platform reach but the institutional capacity to grow the humans who can hold the calls AI cannot make and platforms cannot replace.

That is the operator argument for the next decade. The companies, the careers, and the funds that get this right will compound. The ones that do not will be squeezed from both ends. And the squeeze is already happening.

The middle is where you get paid. Plan accordingly.