Organisation and operations

AI doesn’t replace the team: it multiplies it

Firing someone so another person does the same with AI leaves you where you were. Getting the same team to deliver twice or three times as much puts you ahead.

Reading time: 4 minShare on LinkedIn

AI isn’t there to cut headcount. It’s there so the same team delivers twice or three times as much.

I’m saying this against the current, because the dominant story runs the other way: an agent will do one person’s job, so that person is surplus. The maths behind it is poor. You let someone go, load another person with more volume thanks to AI and, at best, you get the same work you already had, a little cheaper and with a dismissal along the way.

I prefer the opposite maths: the same team, the same salary cost and far more work delivered. That is a competitive advantage.

The slowest junior on the team stopped being slow

A junior in my team took a long time over every task. That’s normal: with new tools, a junior takes much longer than a senior. On top of that, a good part of the work was multidisciplinary and very manual: copying and pasting, moving information from one document to another, layout. Everything got left to the last day and delivered in a rush, and the rush brought small mistakes and typos.

When we brought AI in, we showed them how to automate exactly that part, the part AI handles quite easily. Now AI starts every project: it gathers the details from recorded meetings, email and Asana, takes the person’s suggestions into account and prepares a first draft of the proposal. Then the person sits down to correct it, suggest a different approach and change whatever needs changing.

Within three months, what used to take more than a week was being delivered in a few days, with a much larger volume of work. And the small mistakes disappeared. That’s the part I find most interesting: attention to detail improved the opposite way to the usual one, with less pressure, because AI took over the work that caused the errors.

The same project, before and after: AI takes the manual work and the person keeps the judgement.

If you’re afraid AI will replace you, it will

I say it like that to any team. AI is a tool, and tools have to be learned. Whoever knows how to work with it will have projects, work and responsibilities.

The people who get replaced are the ones who don’t adapt to change: who don’t learn the new tools reaching the market, which are already indispensable. It sounds simple, and it is.

Knowing how to use ChatGPT is no longer enough to work in marketing

This changes who I hire. From now on, every profile I look for will need experience with agentic AI. Having used ChatGPT doesn’t count: almost everyone has, one way or another. I need whoever works in marketing to understand how to automate their work, because I’m also going to ask for considerably more volume than before.

I’ll admit a limitation: I can’t tell you how replacing a person with an agent turns out, because we haven’t done it. And I don’t intend to. I believe people shouldn’t be replaced by AI.

If I were deciding today, I’d do three things:

  • Find the manual work in every role. Copying, pasting, moving information between documents, layout. That’s the first thing AI should take on.
  • Let AI start every project. Have it gather what’s already in meetings, email and tools, and prepare the draft. The person corrects and decides.
  • Hire people who know how to automate. And, just as clearly, raise the volume expected from each person.

Cutting with AI leaves you where you were. Multiplying with AI puts you ahead.