Chesterton’s fence keeps coming to mind as I read yet another article about so-called “TechBros” pushing to replace employees with AI—sooner rather than later.

Below are my thoughts...

AI is getting alarmingly good. Tasks once handled by highly trained professionals—solicitors, radiologists, strategists—are now being taken on by algorithms. So the real question isn’t whether AI can do the job. It’s: should it?

There’s a useful warning from G.K. Chesterton, who said you shouldn’t take down a fence unless you understand why it was put up in the first place. It might look outdated or inefficient, but it could be doing more than you realise.

Today, that fence is skilled human employment. From a purely technical and economic standpoint, replacing people with AI makes perfect sense. AI is fast, tireless, consistent, and scalable. For organisations, that means lower costs, higher output, and a competitive edge. The logic is clear: if a machine can do the job better, why stick with the human?

But the original purpose of the fence isn’t just about ticking boxes or completing tasks.

Jobs aren’t only about productivity. They’re how income is distributed across society. They give people a role, a sense of identity, and a reason to engage. They enable mentorship, skill transfer, and collaboration. Work supports mental health, community, and social stability. Remove that, and you’re not just disrupting business—you’re unravelling parts of society itself.

Pull humans out of the equation too quickly and you risk serious consequences: rising inequality, social disconnection, a loss of purpose, pressure on public services. These aren’t accidental side effects. They’re signs that the fence was doing more than just getting in the way.

And let’s not kid ourselves that AI has nailed everything. Intuition under uncertainty, working through messy or unfamiliar problems, empathy, trust-building, creative synthesis—these remain deeply human abilities. Not every role is rule-based or reducible to data.

So yes, AI is advancing fast. But before dismantling what we’ve built, we need to understand why it was there to begin with. That means asking not only what AI can do, but what human work truly means—and what vanishes if we give it up too readily.

Progress matters. But wisdom lies in knowing which fences are safe to tear down—and which are doing far more than they appear.