Oskar Minge. Every day.

Not long ago, we were impressed that a computer could beat a human at chess. Today, it will write our emails, make a presentation, edit a photo, design a logo, translate a text, suggest a marketing strategy, analyse sales, create music and, with a good enough prompt, probably explain why we've been putting off the dentist for three years.

And it's all wonderful, really. The problem starts when we hand over not just the work, but our own opinions too.

Technology has always helped people do more. The wheel let us carry more, the calculator count faster, the internet find information without going to a library. The smartphone put a camera, map, television, newspaper, shop, bank and a dozen people telling us we absolutely need another bottle of electrolytes in our pocket.

Artificial intelligence is simply the next stage. Except this time, we aren't just automating our hands.

We're automating part of our minds.

And that's where it gets interesting.

I no longer ask whether AI can do something

I work with technology a lot every day. Marketing, graphics, e-commerce, photography, strategy, content creation — in practically every one of these areas, AI can now do something that until recently required a person. Some things it does brilliantly, some averagely, and some with such confidence that you briefly wonder whether you're the idiot. A very human trait, incidentally.

That's why I'm increasingly less interested in asking: “Can AI do this?” Of course it can. A much more interesting question is: “Should I let it do this for me?”

Those aren't the same thing. I can ask AI to write a text, invent an opinion, find arguments. I can even say, “Write down what I should think about this.” And that's precisely where a boundary begins that I wouldn't want to cross.

It's not the machine becoming too clever that scares me

For months, we've heard debates about AI taking people's jobs. It will probably take some. Just as digital photography changed photography, the internet changed newspapers, streaming changed music and the car wasn't particularly good news for carriage makers.

Technology doesn't ask for permission. It simply arrives.

But something else interests me much more. I'm not especially afraid of a world in which AI is very intelligent. I'm more afraid of one in which people stop thinking they need to be intelligent.

Why remember when you can search? Why calculate when you can ask? Why write when software will do it? Why think when an algorithm will provide an answer? And finally: why have an opinion of your own when you can generate the most convincing-sounding one?

To me, that's the truly interesting moment in this technological revolution. Not when machines start doing things like humans, but when humans voluntarily stop doing things that have developed their humanity for years.

You can own a Ferrari and still not know how to drive

AI today is a bit like a ridiculously powerful car. It can get you there much faster. It can also get you into a wall much faster.

Owning a tool doesn't make anyone a specialist. Photoshop didn't make everyone a designer. A smartphone with a good camera didn't make everyone a photographer. Canva didn't create billions of art directors. And ChatGPT won't automatically turn everyone into a strategic genius.

A tool amplifies its user. If you know what you're looking for, it can be an incredible accelerator. If you don't, it may simply produce mistakes much faster than before.

That's one thing I've learned working with AI: the more I know about a subject, the better the results I can get from it. Paradoxically, technology that was supposed to mean we wouldn't need to know everything rewards knowledgeable people even more.

Because they can spot nonsense.

Authenticity will be the greatest luxury

For years, the internet democratised publishing. Everyone got a microphone. AI now gives everyone a copywriter, designer, translator, editor and a small advertising agency living in their phone as well.

The result? An unimaginable amount of content. Correct content. Beautiful content. Professional content. And very often, completely meaningless content.

That's why I feel the greatest value in the years ahead won't be perfection. It will be the person. Character. Mistakes. A particular sense of humour. A sentence that wasn't generated by an algorithm simply because it was the most probable next sentence.

Something real.

I'm not a fan of a world polished so thoroughly that everything looks identical. Brands speak the same language, people use the same phrases, videos have the same transitions and photos the same colours. Then everyone wonders why nobody remembers them.

Perhaps that's exactly why.

AI should take away our boredom, not our personality

I'm not someone who wants to stop technology developing. Quite the opposite. I love tools that shorten the distance between idea and result. If software can do three hours of monotonous work for me — fantastic. I feel no need to romanticise manually copying tables.

But the time we reclaim should go towards more human things: thinking, talking, creating, learning and imagining things that didn't exist before.

Technology should remove friction.

Not the human being.

That's why I never want to reach a point where AI replaces my decisions instead of helping me make them. The difference is enormous.

Perhaps this will be a competitive advantage in a few years

Today, everyone marvels at the speed of generation. I suspect we'll eventually start appreciating something quite different: awareness, selection and the ability to say “no”.

I'm not publishing this, even though I can. I'm not using this idea, even though the algorithm says it should work. I'm not changing my opinion just because five language models gave me five elegantly phrased arguments.

I have my own.

And I'm ready to change it when better facts emerge — not when a better-written paragraph appears.

Perhaps that will one day be a luxury. Not access to information, because we'll have an endless supply. The luxury will be distinguishing information from knowledge, knowledge from wisdom, and your own opinion from text that merely sounds as though someone has actually thought it through.

And that's why I still write

I could, of course, end this piece with a striking AI-generated sentence. It would probably be better stylistically. It would have rhythm. Perhaps even a metaphor.

But that wouldn't feel quite right.

So I'll end it my own way.

I'm not worried about machines becoming increasingly like humans.

I'm more interested in whether humans will start becoming increasingly like machines at the same time.

And I think that's worth keeping an eye on.

Every day.

Oskar Minge