AI Isn't the Bottleneck. You Are.

The wrong people are being blamed

Companies are making people redundant and pointing at AI. The story goes: AI is faster, cheaper, and doesn’t need a salary. So we need fewer humans.

That’s not what’s actually happening.

The real reason people are losing their jobs isn’t that AI replaced their work. It’s that AI created so much new output that leadership couldn’t keep up — and instead of solving that problem, they cut headcount.

That’s not an AI problem. That’s a management problem.

From 5 projects to 50

Before AI, a typical leadership team might oversee five major initiatives at once. Strategy, resourcing, review cycles, decision points — all manageable within the bandwidth of the people at the top.

Now AI can spin up fifty.

Fifty projects. Fifty streams of output. Fifty things that need someone to read them, evaluate them, challenge them, and decide whether they’re worth pursuing.

The AI doesn’t care. It’ll keep producing. But the humans at the top? They’re drowning.

And when executives drown, they don’t hire more reviewers. They cut the people producing the work — because that’s the lever they know how to pull.

The comprehension ceiling

There’s a limit to how much strategic change any organisation can absorb at once. Call it the comprehension ceiling.

Every new project, every pivot, every AI-generated initiative competes for the same finite resource: leadership attention.

• How many pivots can one business comprehend at a time?
• How many strategic directions can a board meaningfully evaluate?
• How many AI-generated reports can a CEO actually read in a week?

AI has no comprehension ceiling. Humans do. And that mismatch is where things break down.

When output exceeds comprehension, one of two things happens:

1. Quality collapses. Projects get rubber-stamped instead of reviewed. Bad work ships. Good work gets buried under volume.
2. People get cut. Leadership can’t process the volume, so they reduce it by reducing the team — mistaking their own bandwidth problem for a headcount problem.

Both are failures of the same thing: not enough human judgement to match AI’s output.

The real constraint on AI

Every conversation about AI focuses on capability. Can it write code? Can it produce strategy? Can it analyse data?

Yes. It can do all of that. Capability is solved.

The unsolved problem is organisational capacity to absorb what AI creates.

It doesn’t matter if AI can generate fifty project plans if your leadership team can only meaningfully evaluate five. The other forty-five are waste — or worse, they’re half-reviewed work that creates risk.

AI’s constraint isn’t intelligence. It’s the number of skilled humans available to:

• Evaluate whether the output is actually good
• Prioritise which outputs deserve investment
• Decide when to double down and when to kill something
• Integrate AI-generated work into real-world strategy
• Absorb the organisational change that comes with acting on it

These are leadership functions. They don’t scale with compute. They scale with people.

Why the redundancies are backwards

Here’s the irony: companies are cutting the exact people they need more of.

They’re removing mid-level managers, senior reviewers, and experienced practitioners — the people whose judgement is required to make AI output useful.

Then they wonder why quality drops. Why projects stall. Why AI adoption “isn’t delivering the ROI we expected.”

It’s not delivering because you fired the people whose job it was to turn AI output into business outcomes.

AI produces. Humans decide. If you cut the deciders, you don’t get efficiency. You get chaos with a faster engine.

What should actually happen

Organisations that get this right will do something counterintuitive: they’ll hire into the review layer.

Not more prompt engineers. Not more AI specialists. More people with domain expertise, strategic judgement, and the experience to know what good looks like.

The org chart needs to flip. Instead of a wide base of producers and a narrow band of reviewers, AI-native organisations need:

• Thin production layer — AI handles most of the raw output
• Thick review layer — skilled humans evaluate, refine, and prioritise
• Focused leadership — executives who are ruthless about what gets attention and what gets killed

The companies that win won’t be the ones that produce the most with AI. They’ll be the ones that can comprehend the most.

The question nobody’s asking

Everyone asks: “What can AI do?”

The better question is: “What can your organisation actually absorb?”

Because AI can do almost anything now. The limit isn’t the machine. The limit is you — your attention, your judgement, your capacity to decide.

And until leaders acknowledge that the bottleneck has moved from production to comprehension, they’ll keep making the same mistake: cutting humans to solve a problem that only humans can fix.

AI isn’t the bottleneck.

You are.