The Imaginary Battle

There is a particular moment in getting older that nobody warns you about.

It isn’t the grey hair. Hair has never been especially good at keeping secrets anyway. It isn’t the wrinkles, the slower body, the increasingly creative noises produced when you get out of a chair. Those are merely negotiations with physics.

The frightening moment is much quieter.

You discover that something new has happened, and instead of being fascinated by it, you feel tired. That is when you begin to suspect that perhaps you have become old.

I had this suspicion recently, and it was caused by artificial intelligence.

Which is rather insulting.

I have always been enormously curious. Not usefully curious, necessarily. My curiosity has never shown much respect for practicality. If somebody mentions something I have never heard of, my immediate reaction is not, “How can this benefit me?” It is, “What is that?”

Then come the questions.

What is it?

Why does it exist?

Who thought of it?

How did it begin?

What happened before it?

What happened afterwards?

Does it still exist?

What does it have to do with that other thing?

And, inevitably, some completely unreasonable question nobody involved had anticipated.

…I have occasionally suspected that conversations with me must feel like being interrogated by an unusually enthusiastic customs officer…

But this curiosity has been one of the great pleasures of my life. I have never needed a subject to be useful before becoming interested in it. The world has always seemed too large to restrict one’s attention to things that might eventually appear on an invoice.

Then came AI.

Suddenly I could ask almost anything.

And I did.

It was intoxicating.

A question that once required a library, a knowledgeable friend, an expert, a journey, an afternoon of searching, or several years of education could suddenly be asked in one sentence. AI became a kind of search engine on steroids.

But then I began seeing something else.People were teaching themselves “AI skills.”

People were learning elaborate prompting techniques.

People were writing enormous prompts containing instructions, roles, constraints, procedures, frameworks, desired tones, forbidden behaviours, output structures and what sometimes appeared to be the complete constitution of a small republic. Seven pages of instructions to ask a machine a question.

And I found myself thinking:

**Why?**

That question bothered me more than it should have. Perhaps because I was beginning to feel that I had somehow fallen behind.

Other people were learning how to master the new technology. They were developing sophisticated workflows. They were teaching Claude to do this and ChatGPT to do that. They were discussing prompt engineering as though they had discovered fire.

And I was sitting there thinking:

“I just want to know why this thing happens.”

That was when the alarming thought arrived.

Perhaps I was becoming old. Not physically old. Something worse.

Intellectually old.

Perhaps I was beginning to lose interest in progress.

Perhaps the world was changing faster than I wanted to follow.

Perhaps I had reached the point where new technology was no longer an invitation but an inconvenience.

The thought was disturbing because curiosity has always been part of my identity. If curiosity goes, what remains?

A person can survive without knowing the latest technology.

A person can survive without understanding every new platform, device, software system or fashionable methodology.

But what happens when a person stops wanting to know?

That is different. That is not ageing of the body. That is ageing of the mind.

And here is where AI becomes genuinely interesting. Because AI has not merely given us more information; it has changed the “cost of curiosity”

For most of human history, curiosity was expensive. You wanted to know something, so you had to go looking. You needed books.

Teachers.

Libraries.

Experiments.

Travel.

Correspondence.

Conversation.

Time.

Patience.

Sometimes money.

Sometimes an entire lifetime.

Even finding out whether a question was worth asking could require considerable effort. AI removes much of that friction.

Ask.

Receive.

Next question.

It is an extraordinary gift to a curious person, and also a potential disaster.

Because friction has a strange relationship with thought. Sometimes friction is what makes us think. When information is difficult to obtain, we remember the journey. We compare sources. We notice contradictions. We develop judgement. We become invested in the subject. When information becomes effortless, something dangerous becomes possible:

“We can consume answers without becoming interested in the thing itself.”

The distinction is subtle.

I can ask why medieval cities were built where they were and receive a perfectly competent explanation. I now know something. But knowing the answer isn’t necessarily the same as understanding the subject. Understanding begins when the answer produces another question.

Why there?

What would have happened if the river had moved?

Why did one city flourish while another disappeared?

How much of this was geography and how much was political power?

Does the same principle exist in modern cities?

That is curiosity. The first answer is information. The second question is thinking.

And this is where I think the current obsession with “AI skills” becomes slightly absurd.

There are certainly situations where sophisticated prompting is useful. If someone is building complicated workflows, programming, analysing large amounts of information, generating structured documents or integrating AI into professional systems, then yes, learning how to communicate precisely with a machine is sensible.

But somewhere along the way, a useful skill has been transformed into a performance.

People are beginning to talk about prompting as though the machine is an extremely temperamental genie who will grant your wishes only if you phrase them according to ancient ceremonial law.

You must first establish the role.

You must specify the objective.

You must provide context.

You must define the constraints.

You must specify the desired output.

All sensible.

Until someone produces a prompt longer than the question they actually wanted answered.

At that point, something has gone wrong. The machine was supposed to reduce friction. We have built a new kind of friction around the machine. People do this remarkably well.

We invent a tool to simplify something, then invent an industry explaining how to use the tool properly, then create courses teaching people how to become experts in the industry created around the tool.

Eventually someone sells a certificate.

Civilisation continues.

The deeper issue, however, is not prompting.

It is the control; people are understandably anxious about losing control to AI.

But I think there is another kind of control we should worry about losing. The control over “what deserves our attention.”

AI can tell us almost anything. That does not mean everything deserves to be known. Nor does it mean everything we can ask should be answered immediately. Curiosity requires selection. A mind cannot investigate everything. Even an extraordinarily curious mind has limited time.

The problem therefore isn’t that AI can answer too many questions.

The problem is that it can make us forget that we are the ones who choose the questions.

That distinction matters enormously.

AI can explain quantum mechanics.

It can explain ancient mythology.

It can explain monetary policy.

It can explain the history of buttons.

It can explain why people fall in love with fictional characters.

It can explain why uniforms make people look more attractive.

It can explain the psychology of jealousy.

It can explain why we wake up at three in the morning and stare at the ceiling as though the ceiling has personally betrayed us.

But AI didn’t decide that these things were interesting.

We did.

The machine can follow curiosity, but it cannot replace the peculiar human impulse that says:

“Wait. I want to know that.”

And perhaps that is the part of ourselves we should protect most carefully. Because there is a genuine danger in artificial intelligence, but I don’t think it is the one usually presented. The danger isn’t necessarily that machines will become smarter than us. They already outperform us in plenty of narrow tasks.

The danger isn’t even necessarily that machines will replace human labour. They certainly will replace some kinds of labour, and pretending otherwise is merely comforting ourselves with optimistic wallpaper.

The deeper danger is the premature satisfaction.

We ask.

The machine answers.

We move on.

Question answered.

Curiosity extinguished.

Next.

That is where intellectual laziness can hide inside technological efficiency. A person may become extraordinarily efficient at acquiring information while becoming increasingly incapable of dwelling with an idea.

And dwelling matters. Some ideas need time. Some questions should remain unanswered for a while. Some subjects become interesting only after the first obvious explanation has been exhausted.

There is a great intellectual difference between saying:

“I know the answer.”

and saying:

“I understand why that answer might be true, what it leaves unexplained, what could contradict it, and what question it leads to next.”

AI is exceptionally good at the first.

It can also help enormously with the second.

But it cannot make us do it. That remains our responsibility.

Which brings me back to my imaginary battle with AI.

I thought I was losing. Other people seemed to be running ahead, acquiring skills, mastering tools, developing elaborate systems.

I wasn’t.

I was asking questions.

And suddenly I realised that perhaps I had misunderstood the contest.

There was never a contest.

I don’t need to beat AI.

I don’t need to become better at producing answers than a machine.

I don’t need to compete with something whose entire existence is devoted to processing information.

That would be like challenging a calculator to arithmetic and then becoming depressed because it won.

The interesting question is not:

Can I do what AI does?

The interesting question is:

What can I do with what AI can do?

That changes everything.

If AI can retrieve information in seconds, I can spend less time retrieving information and more time examining it.

If AI can explain something, I can challenge the explanation.

If AI can produce an argument, I can attack it.

If AI can generate ten interpretations, I can decide which one seems convincing.

If AI can connect two distant subjects, I can follow the connection somewhere neither of us expected.

And if it gives me an answer that feels too neat, I can say the most useful word in the human language:

But.

But why?

But, is that really true?

But, what about this?

But, doesn’t that contradict the other thing?

But, what happens if we reverse it?

But what does this mean?

That little word is a surprisingly powerful defence against intellectual passivity. It keeps the conversation alive. It prevents an answer from becoming an ending.

And perhaps that is the real skill we need in the age of AI.

Not prompt engineering.

Not AI fluency.

Not knowing which incantation produces the most polished paragraph.

Question fluency.

The ability to recognise an interesting question.

The courage to ask a stupid one.

The patience to pursue an apparently useless one.

The scepticism to challenge a convincing answer.

The imagination to ask what follows.

And the discipline to notice when the machine is giving us something that sounds intelligent without actually being right.

These are not obsolete skills. They may become more important precisely because information is becoming abundant.

When information was scarce, knowing things was valuable.

When information becomes abundant, knowing what to do with what you know becomes valuable.

And when answers become abundant, perhaps the most precious thing becomes the ability to remain dissatisfied with them.

That is not dissatisfaction in the miserable sense. It is intellectual appetite. It is the refusal to let the first answer close the door.

Perhaps, then, growing older does not mean becoming less interested in progress.

Perhaps it means learning to distinguish progress from novelty.

Every new technology is not automatically progress.

Every new method is not automatically an improvement.

Every new “skill” is not worth acquiring simply because somebody has built a business around teaching it.

Growing older may actually give us an advantage here.

We have seen fashions come and go.

We have watched technologies arrive promising to transform everything.

We have watched people become ecstatic about things that disappeared five years later.

We have learned, sometimes painfully, that novelty and importance are not synonyms.

That isn’t intellectual decline; it is discrimination.

The real warning sign would be something else.

It would be encountering something unfamiliar and feeling no desire to understand it.

It would be hearing a strange word and not wondering what it means.

It would be meeting an unfamiliar idea and deciding immediately that it is irrelevant.

It would be losing the itch.

That is the ageing I fear, not the body becoming older.

The mind becoming ‘finished*’.

Because curiosity keeps the future open.

A curious person is never entirely settled inside the world they already understand. There is always something outside the circle of knowledge, something irritatingly unexplained, something that refuses to fit, something that suddenly appears and demands attention.

AI has made that circle strange. It has brought an enormous portion of human knowledge within conversational reach.

We can now ask questions at a speed previous generations could scarcely imagine. That is not the death of curiosity, it is an unprecedented opportunity for it.

But only if we remember the difference between “having an answer and having a question.”

The answer closes.

The question opens.

And perhaps the greatest irony of all is that the technology that made me briefly wonder whether I was becoming old has reminded me of exactly why I am not.

I am still asking.

I am still irritated by things I don’t understand.

I am still following irrelevant questions down completely unnecessary intellectual corridors.

I am still capable of encountering something I have never heard of and thinking:

What the hell is that?

Then I want to know.

And then I want to know what the answer means.

And then, inevitably:

But…

So, perhaps I haven’t lost the battle with AI after all.

There was never an imaginary battle. There is only a much older one, the one human beings have been fighting since they first looked at the world and refused to accept that not knowing was good enough. The machine can give us answers faster than anything we have ever built. It can search farther, connect more information, explain more subjects and generate more possibilities than any individual human could hope to master.

Let it.

That is what tools are for.

Our job is not to become better machines.

Our job is to remain the creatures who look at the answer and ask:

“Yes. But what does it mean?”