← Back to archive

Why AI Companies Will Have Fewer Managers and More Agents

The future company has fewer managers/specialists/SaaS tools .... and FAR MORE agents.

What I’ve noticed is that more and more companies reaching out to us repeat the same pattern. They feel behind - they can see the world moving and they know their own operation is not moving at the same speed.

There is a lot to optimize inside those companies. A large part of what people do all day is monkey job: moving information from one place to another, reformatting it, waiting on someone for an answer. Sound familiar? Agents do that fantastically well.

And yes .... that means layoffs.

There is no soft way to say it. When the work disappears, the role goes with it. I expect more companies overall, and maybe more jobs in total, but fewer people inside each one. The company does not need to carry every function as a full department.

I do think companies can do the same with fewer people, and in some cases more.

The shift is structural. AI does not only make the people you already have faster. It changes the optimal shape of the company: fewer people, less hierarchy, more internal software, and humans supervising fleets of agents.

Roles are collapsing: design, engineering, product management overlap more every year. Instead of a designer handing over a static mockup, one person can build the working prototype.

Engineering moves the same way. Less time typing, more time describing the problem properly, then checking what comes back. The mechanical part is the part that goes.

Fewer SaaS tools follow from the same thing. If someone can build the internal tool that connects exactly the data they need, and agents can maintain and secure it afterwards, most of the reasons to buy that tool are gone. You stop shaping the company around Salesforce and start generating software around how you actually work. We literally saved hundreds of hours of manual work by building a custom tool in less than a couple of days.

More and more of the CEO job is routing. Connecting people, context, decisions. That kind of coordination can be automated, and what is left for humans is invention and judgment.

Removing the bureaucracy and the repetitive work matters more than squeezing ten times the output out of whoever stays. A smaller company waits less: fewer status loops, fewer meetings that exist only because nobody can see the whole system. The hours that remain should go into something worth doing.

A robot contrasts with a human, listing humorous reasons why AI can't replace human qualities and experiences.

Originally posted on X

PS: I'm pretty active on 𝕏 if you'd like to follow along

Archive

2026

2025

2024

2018