AI adoption is being framed as a headcount reduction story

Originally published on LinkedIn

AI adoption is being framed as a headcount reduction story.

I think that misses the bigger picture.

If AI increases output capacity, leadership ambition should increase with it.

The question should not only be “how many people do we still need?”

It should be: “what can this team now build, automate, test, launch, or improve that was previously impossible?”

Coinbase announced it is laying off 14% of its workforce as part of a push to become leaner, faster, and more AI native.

I understand the logic. Coinbase wants to flatten the organization, reduce coordination tax, and become more AI native.

The thesis seems to be that managers now need to be both managers and ICs.

I’m not against player-coach managers.

I’ve worked that way for years. In smaller, faster teams, the best people often need to manage, build, write, analyze, automate, and make decisions.

That part makes sense.

They’re also advocating for this “one-person team” concept, with the roles of engineer, designer, and PM all rolled up into one.

Personally, I’m a big fan of that, although I don’t believe it’s for everyone.

A one-person team can be a powerful operating model.

But it becomes dangerous when it means doing more with fewer people.

The focus shouldn’t be on the quantity of people but on the quality of the output.

If Coinbase is well capitalized, if crypto is on the verge of the next adoption wave, and if AI-powered ICs are going to be much more productive and able to ship faster, why the layoff?

Why let go of productive employees who already know the industry, the customers, the product, and the internal systems?

Why not equip the existing workforce with the best AI tools and generate more output, better output, or entirely new kinds of output?

Of course some roles will disappear.

Of course some teams are bloated. Of course AI changes the cost structure.

But layoffs as the first move reveal a lack of imagination.

Before using AI as a headcount reduction lever, companies should redesign the work.

Map the workflows. Identify where AI expands capacity.

Train the people who already understand the business.

Redeploy them into higher leverage work.

Then, and only then, decide which roles are genuinely obsolete.

Otherwise, we are not adopting AI. We are just using AI as cover for the same old cost-cutting playbook.