The Floor.

There is something the AI consulting industry does not want to say, so most of us dance around it. I want to say it plainly.

For some roles, AI is not an augmentation tool. It is a replacement. The work the AI does is the same work the human does, and the AI does it faster, cheaper, and increasingly better. There is no "higher-value frontier" for the human to move to within that function. The role is going away.

This is not true for most roles. It is not true for the majority of what organizations do. But it is true for some roles, and pretending otherwise is what produces the worst outcomes.

The dishonesty problem

The standard narrative goes like this: AI will not replace people, it will augment them. Everyone will move to higher-value work. We just need to upskill, reskill, and redesign roles. The future is human-AI collaboration.

This is true for many roles. For Principal roles, where the human owns the outcome and the AI is their instrument, it is exactly right. For Architect roles, where the human designs systems of agents, it is aspirational but possible. But for a meaningful subset of roles - the ones where the human is structurally positioned as a Competitor to the AI - the collaboration narrative is a polite fiction.

And people know it. They can feel the distance between what they are being told and what they observe. The training program says "augmentation." The headcount plan says something else. That gap between rhetoric and reality is where trust dies.

When the organization says "nobody is losing their job to AI" and people can see that is not entirely true, the credibility loss extends far beyond the affected roles. It poisons adoption everywhere.

What happens when you pretend

I have seen this pattern repeat across organizations. Leadership avoids naming the floor - the roles where displacement is the honest outcome - and instead wraps everything in the augmentation narrative. Here is what follows.

People in the affected roles are not fooled. They know their work is being automated. But because the organization will not name it, they cannot plan. They cannot negotiate. They cannot even grieve. They are stuck performing cooperation while quietly preparing to leave or disengage.

People in adjacent roles watch closely. They are not in the displacement zone, but they are near it. How the organization treats the affected roles is how they expect to be treated. If the floor is handled dishonestly, everyone adjusts their cooperation downward. "If they lied to that team, they will lie to us."

Champions burn out. The people who invested the most in making AI work - who trained others, who redesigned workflows, who advocated for the tool - watch their contributions lead to their colleagues being quietly managed out. The reward for championing AI was enabling someone else's exit. They stop championing.

What honesty looks like

Naming the floor does not mean being callous about it. It means being precise. The diagnostic should produce a clear answer for each function: is the human role here a Competitor, a Principal, or an Architect? And for Competitor roles where there is no realistic path to Principal, the honest assessment is: this role is going to shrink or disappear. The question is how.

The key insight: the honest approach is not just more ethical. It produces better AI outcomes. The leavers cooperate because the terms make cooperation their best option. The survivors cooperate because they saw the affected people treated with respect. The champions sustain because their work led to a fair process, not a betrayal.

The transition fairness lever

For roles at the floor, the only lever that produces genuine cooperation is transition fairness. This is not a euphemism for a better severance package. It is a specific design:

Most consultancies will not name the floor because it is uncomfortable and it complicates the sales conversation. It is easier to sell "AI augments everyone" than to say "AI augments most roles but displaces some, and here is how to handle both."

But naming the floor is what makes the rest of the framework credible. If you are honest about the roles where cooperation is not rational, your prescriptions for the roles where cooperation is achievable become trustworthy. The organization knows you are not selling a fantasy. And that trust is what makes the harder work - redesigning roles, rebuilding incentives, measuring what matters - actually possible.

The floor exists whether you name it or not. The only choice is whether you address it with a plan or let it address you through attrition, distrust, and failed adoption.