Are we outsourcing the very faculty that built AI in the first place?

There was a time, not so very long ago, when intelligence was not artificial. It was simply a human being, thinking.

We solved problems by watching, reasoning, remembering, and imagining our way toward an answer. We argued and debated. We questioned, experimented, failed, and went back to try again. The process was slow and often maddening, but it produced something a machine still cannot hand us: judgment earned the hard way. Whether we turned out right or wrong almost didn’t matter. The understanding was ours, because the effort had been ours.

Today we stand on a somewhat unfamiliar surrounding.

For the first time in our history, we have built machines that can write an essay, draft a contract, compose music, generate a photograph, debug code, and hold a conversation convincing enough to forget there is no one on the other side. I belong to the cusp generation — old enough to remember the other way of doing things, young enough to use the new one every day. I can still picture myself at the office late in the evening, drafting by hand a felicitation note for the company’s anniversary, hunting for a slightly different line for each person so that no two people received the same congratulation. A single festival greeting could take the better part of a night. It seems almost quaint now. A few keystrokes and the machine offers me fifty versions before I’ve finished my tea.

Artificial intelligence is, without exaggeration, among the finest things our species has ever made.

And yet a question keeps returning, and I cannot quite talk myself out of it.

What happens when we lean on artificial intelligence so completely that we stop exercising the real thing?

In small ways, it has already begun. There was a time most of us could run a column of figures in our heads. Then the calculator arrived, and within a generation the skill quietly left us. You can still find it here and there — the older shopkeeper who glances at your basket, pauses for a beat, and tells you the exact total before you’ve reached for your phone. But he is the exception now. Take away most people’s calculator and watch them stall over a sum a schoolboy once managed standing up.

The irony at the heart of it

The irony is too large to step around.

Artificial intelligence did not descend from some machine. It came out of us. Every algorithm, every neural network, every stubborn line of code began in a human mind — people who spent decades thinking hard, asking awkward questions, and sitting with problems that refused to yield. Human intelligence is the parent of artificial intelligence.

What unsettles me is how quickly the child is being asked to replace the parent.

The student hands the assignment to it. The professional lets it write the report. The manager asks it to summarise the meeting he didn’t fully attend. The writer asks it for the idea he used to chase down himself. And in courtrooms in more than one country, judges have begun consulting it on the very questions they were appointed to weigh. A tool meant to extend human capacity is, by degrees, becoming a substitute for human effort.

Every convenience sends a bill

Each leap forward arrives wrapped in convenience. The calculator spared us the arithmetic. GPS spared us the map. The search engine spared us the trouble of remembering. We made those trades and, on balance, were richer for them.

This one is different in kind.

The earlier tools took over tasks. This one offers to take over the thinking itself — and that is a wholly different bargain.

A muscle left unused does not stay strong; it softens and forgets its own strength. The mind is no different. The real risk was never that the machine would out-think us. The risk is duller and closer to home: that we will simply lose the appetite to think for ourselves.

The lost art of the struggle

Almost everything I have learned that was worth keeping, I learned the slow way.

A child learns by falling over. A founder learns by getting it wrong with his own money on the line — I spent the better part of three decades building a company up from almost nothing, and not one useful lesson in all that time arrived gift-wrapped. A leader learns in the cold minutes when a hard decision has only him to make it. The struggle was never the obstacle standing between us and the lesson. The struggle *was* the lesson.

When every answer appears the instant we want it, we quietly give up the journey that turns information into understanding. Being handed an answer and arriving at one are not the same act. One leaves you with a fact. The other leaves you changed.

What we are handing the young

My real worry sits with the generation coming up behind us.

Many of them are growing up in a world where the answer arrives before the question has properly formed. It began innocently enough with Google and has become Gemini, ChatGPT, and whatever you happen to subscribe to next — and the design is the same each time. The first taste is free and frictionless, generous to a fault, right up until the dependence is complete. We have a word for that pattern, and it isn’t *convenience*.

Why spend an afternoon in the research when a summary lands in seconds? Why build an argument when one can be conjured whole? Why labour over a paragraph when the machine writes faster than you can think?

The technology is not the villain here. The danger is the moment convenience quietly replaces curiosity. Every real advance in human history came from someone who refused the accepted answer, distrusted the obvious, and went looking for what did not yet exist. If the next generation grows up consuming intelligence rather than producing it, I wonder what becomes of invention itself.

A bicycle, not a chauffeur

Good tools leave you stronger for having used them.

Someone once called the computer a bicycle for the mind, and it remains the right picture. A bicycle carries you further and faster than your legs alone ever could — but it still asks you to pedal, and you arrive fitter than you left. The danger today is that we are quietly trading the bicycle for a chauffeured car: it gets us there in comfort, asks nothing of us, and teaches us nothing about the road.

In the Army they taught us, in a hundred uncomfortable ways, that an outfit which stops training does not stay sharp for long. Capability is not a possession you bank. It is a discipline you keep — or slowly lose.

Where the real future lies

None of this is an argument for switching the machine off. That would be both foolish and impossible, and I have no interest in it.

The future will belong to people who can hold human judgment and artificial capability in the same hand. Let the machine do what it is extraordinary at — sift oceans of information, throw up a hundred possibilities in a minute, draft the rough first version. Then do what it cannot. Supply the context. Decide what any of it means. Choose, out of all those possibilities, which question was even worth asking.

The most valuable skill of the next twenty years may not be knowing how to use artificial intelligence at all. Almost everyone will manage that. The rarer thing will be the ability to keep thinking for yourself while you use it.

One question, before we get too comfortable

So perhaps, in the rush to embrace something genuinely remarkable, it is worth stopping long enough to ask ourselves a single plain question.

Are we using artificial intelligence to enlarge our intelligence — or to retire it?

The answer will shape more than the future of work. It will shape what we remain capable of.

Because long before intelligence was ever artificial, it was deeply, stubbornly human. And it was that intelligence — ours — that built the world we are now so tempted to hand over.