Select Page

Something has been quietly taken from our people over decades, and I think AI, oddly, is giving it back.

Watch the arc. Industrial-era process design made each of us a cog: responsible for a step, blind to the whole. Then enterprise software and SaaS arrived, and the logic of our work moved into the platform. The system decided the sequence, enforced the rules, embedded someone else’s idea of best practice. Deterministic and dependable, but the thinking had been done elsewhere. People stopped owning outcomes and started following the bouncing ball. Ask someone to describe their process today and many can’t. They do stuff. The screens tell them what’s next.

That’s the slow erosion of agency, and organisations barely noticed because the work still got done.

Then agents arrived, and the game changed. To get real value from an agent you must be able to say what the work actually is. What activities happen, why, feeding what outcome, judged by what standard. The agent has no bouncing ball to follow; you have to articulate the intent the SaaS era let everyone forget. And that requirement radiates in both directions. Downstream: does this process have an owner, how is success measured, is anyone actively managing it? Upstream, and this is the question I love: why does this activity exist at all, and what value stream does it serve? Many can’t answer, not through any fault of their own; the machine had absorbed the why.

So the very technology people fear will strip their agency actually demands its return. Agents force organisations to hand understanding, intent and judgement back to humans, because without those, the agent is useless. The person who deeply knows why the work matters becomes more valuable, not less. The person who only knew which buttons to press was already automated; it just wasn’t visible yet.

Ask your people to describe their process, its purpose, and what good looks like, without mentioning a system. Where they can’t, that’s not their failing. It’s a map of everywhere your organisation outsourced its thinking, and everywhere AI will hand it back.