Where will experience come from?

We want people with experience. I’m wondering where we expect them to get it if we automate the work they used to learn on.

At SWITCH in Singapore this week, the announcement of the first Champions of AI companies put workforce development and redesigning roles alongside business transformation. That feels like the right conversation to be having. What changes for the people doing the work? Source

Earlier this month, UOB offered a useful example. As AI reshapes entry-level work, the bank said it expects to recruit 1,000 young graduates over a year, across banking roles and early-career programmes. It’s also investing in developing its managers for an AI-enabled workplace. Source

Those are plans, and the results will take time. But I like the willingness to keep investing in people while the work itself is changing.

Some of the tasks we now want AI to handle were also how people learnt the business. Preparing a first proposal. Checking a report. Following up when something didn’t add up. Sitting with someone more experienced who could explain why the apparently obvious answer wouldn’t work.

Not all of that work was valuable. There’s no virtue in making someone spend three hours formatting a spreadsheet just because we once had to. But some of it gave people a chance to build judgement, make a manageable mistake and understand what happened next.

If we’re taking those tasks away, what are we putting in their place?

I’ve been thinking about this through my mentoring with SDG Open Hack 2026 here in Singapore, supporting students as they develop ideas around real sustainability challenges.

My role is often to ask fairly simple questions. Who would use this? What problem does it solve for them? What would you need to test before taking it further?

The value isn’t in giving a team my answer. It’s in giving them room to work through their own, question an assumption and decide what to do next. That takes practice. It also takes someone being willing to spend time with them.

And a very big congratulations to HydraCool, the team I’ve been mentoring, on winning the Grand Finale! I’m delighted for them. The achievement is theirs, and it’s been a pleasure to support them along the way.

A student challenge and a workplace are different settings, of course. But both need opportunities for people to take responsibility, explain their thinking and learn from feedback.

For business leaders, I think this belongs in the AI transformation plan from the start. Alongside the time saved and the cost reduced, we should be asking how the next generation will learn to run the business.

Otherwise, we risk asking for experienced people while quietly removing the opportunities that help create them.

I’d start with three questions:

• What did the work we’re automating teach someone?
• Where will they get that experience now?
• Who has the time and responsibility to help them develop?

That could mean letting a junior colleague lead a customer conversation with support, work through an unusual case, or explain why they’d accept or reject an AI recommendation. Real responsibility, with someone available to help.

It also means giving managers time to do it. Mentoring won’t happen consistently if everyone’s measured only on this week’s output.

I’m all for using AI to make work better. I’d like some of the time it gives us back to go into helping people get better at it too.