Roles have always changed. Professions have always disappeared. New ones have always taken their place. The weavers who smashed the looms weren't wrong to be afraid — they were right about what was happening to their world. They were wrong about what to do about it. But the fear was rational. The disruption was real. Their livelihoods were ending faster than any reasonable person could adapt.
That's the frame that most AI commentary avoids. It's easier to talk about opportunity, about capability, about the jobs that will be created. All of that is true. None of it is the full picture.
The roles most immediately affected by AI are not always the ones most discussed. Junior developers are a common example. But the more immediate pressure is elsewhere — on the person producing mockups, on the analyst generating weekly reports, on the coordinator managing handoffs between teams.
These roles aren't disappearing because AI is smarter. They're changing because AI now handles the mechanical part of the work — which means the human part has to be different. More judgment. More interpretation. More of the things that were always there but were buried under volume.
At the same time, other roles become more valuable — not less. The architect who automates build pipelines. The designer who translates process into experience rather than producing screens. The programme manager who understands what AI can and cannot reliably do in a delivery context. The shift is not uniform. And treating it as if it were produces strategies that are right in aggregate and wrong for almost everyone affected.
Most organisations approach this in one of two ways. The first is dismissal — AI is a future problem, something to revisit when it becomes unavoidable. The second is anxiety — every conversation about AI carries an unspoken subtext of job loss, which makes honest discussion nearly impossible.
Neither is useful. Both are understandable. And both are, in their own way, a form of simplification — a way of reducing a complex, uncomfortable reality to something that feels manageable.
The alternative is harder. It requires holding the full picture — the opportunity and the disruption together, without resolving the tension prematurely in either direction. It requires being honest about what is changing, for whom, and at what pace — before building any kind of response to it.
Start with an honest inventory — not of headcount, but of work. What does each role actually consist of, at the level of specific tasks? Which of those tasks are candidates for assistance? For automation? For zero-touch handling? And which require more human judgment than before, precisely because the mechanical work around them is gone?
This inventory doesn't live at the leadership level. It lives with the people doing the work — and most importantly, with the team leads who are close enough to both to see what's actually changing and what it means for the people they're responsible for.
The organisations that get this right don't cascade a strategy downward. They build the picture from where the work happens — and they give team leads the space and the framework to develop that picture for their own teams. Not to soften the message. To get a more accurate one.
People can handle change when they can see where they're going. What they cannot handle is being told the destination without being given any say in how to get there — or any honest account of what gets lost along the way.
The Luddites weren't irrational. They were unheard. The response to their disruption was suppression, not support. The lesson wasn't that technology should slow down. It was that transitions require something more than a roadmap and a change management workstream.
They require the willingness to hold the complexity — the opportunity and the loss, the potential and the cost — without simplifying it away before the people most affected have had a chance to be part of the answer.
That's not a communication strategy. It's a starting point. And it's the one most organisations skip.