Will AI replace my team? What really changes in the daily routine

In a well-designed system, AI takes the repetitive part of the work off your team: copying data, confirming, sorting messages, putting together the first draft. Decisions, exceptions and contact that calls for judgment stay with people. What changes is where the team spends its attention. And the people who know the process best, your team, need to take part in the design.

She has been with the company for eight years. She knows which client likes to be called by a nickname, which supplier is always late and why that spreadsheet has a hidden tab. When you mention “implementing AI,” she goes quiet in the meeting.

You know what she’s thinking. It’s worth answering honestly, for her and for you.

What AI does well

Tasks with volume, a pattern and little variation. Reading a messy message and understanding it’s a rescheduling request. Summarizing a long conversation. Sorting requests by subject. Preparing a draft reply for someone to review.

This is real work, and it takes up a big chunk of a lot of people’s day. It’s also the chunk almost nobody enjoys.

What stays with people

The philosopher Michael Polanyi wrote in 1966 that we know more than we can tell. That’s exactly what the eight-year employee carries: knowledge that was never written down because nobody needed to write it. No system learns that on its own. It only works if someone translates part of that knowledge into rules, examples and limits.

And then there are the exceptions. The angry client, the out-of-the-ordinary request, the situation no rule anticipated. In 1983, the researcher Lisanne Bainbridge published a paper called Ironies of Automation. One of the ironies she points out: the more the system takes over the routine, the more important the person who handles what’s left becomes, because what’s left is exactly the hard part.

Automating well raises the value of whoever knows how to handle exceptions.

The uncomfortable part

It would be dishonest to say nothing changes. It does. If an entire role comes down to copying information from one system to another, that task is going away, with or without AI. A simple integration already does it.

The useful question for whoever makes the decisions is a different one: what would this person do with ten extra hours a week? At most service businesses, the answer comes fast. Get back to the client who was left waiting. Follow up on the difficult case. Do the thing that always gets pushed to later.

If the answer is “nothing,” the problem started before the technology. And it’s the leader’s job to fix that, with or without a new system.

How to talk with your team

• Bring the people who do the work into the mapping. Nobody knows better where the workflow gets stuck.
• Say in writing what changes and what doesn’t.
• Show, at each stage, where the person still makes the call.
• Start with a small pilot and ask the team to evaluate it.
• Agree that if the system gets in the way, it gets adjusted or turned off.

A team that takes part in the design runs the system. A team caught by surprise learns to work around it.

If you want to see which tasks in your operation can move off your team’s plate, and which should stay, the AI-ization Diagnostic shows you in a few minutes.

What would your team do with a free Friday afternoon?