If AI doesn’t take over the world, we’ll continue using it to build great — and terrible — software.

Over the past few years, the barrier of entry has collapsed. For roughly the price of a $20 subscription, you can design, build, write, animate, ship and market something that would once have required an entire collection of specialised skills, months of time investment and the combined effort of experts. Increasingly, you can do all of this without ever learning how the underlying systems work. Touching the code is optional. Understanding the infrastructure is optional. Sometimes, even leaving the chat window is optional.

That doesn’t mean expertise has become worthless. Far from it. It means that production itself is no longer the bottleneck it once was.

The important question is shifting from can you build this? to should you build this?
And, if you do:
Why should anybody care?

Finding useful information or helpful software used to be the problem. Now, there is so much being produced, so quickly, that the difficult part is filtering through the wasteland of AI slop to find the things that feel human.

Things with taste.

Everybody understands the feeling, even if they don’t call it that. You visit two websites that do roughly the same thing. Both work, and both have the features you need.

But one just feels better.

The spacing feels right. The typography feels deliberate. The language sounds human. Buttons are where you expect them to be. Animations happen when they should. The product seems to understand what you’re trying to do before you have to think about it.

The other one technically works, but something feels slightly off.

Maybe there’s too much going on. Maybe everything is fighting for your attention. Maybe the writing feels artificially and generically generated. Maybe the product has fourteen features when it needed four, and sometimes, all of the above and more.

Both were built competently.

The difference is taste.

Taste is what people come back for, and what they are willing to spend their hard-earned money on.

I used to think of taste mostly as aesthetics: something elegant, modern and premium. I still think those things matter, but that definition feels incomplete.

Taste is closer to having a coherent, developed and distinct point of view.

It is knowing what should be there and, perhaps more importantly, what shouldn’t. It is understanding the person you are building for. It is restraint. Hierarchy. Language. Usability. Timing. Proportion. It is hundreds of tiny decisions that individually seem insignificant but collectively determine how something feels.

And as the number of software products explodes, I keep coming back to the same question:
If anybody can make this, what becomes the differentiator?

What separates one fitness tracker from another? Why does one succeed where a hundred others never see the light of day?

Why do people become almost emotionally attached to one notes app when a dozen competitors can store the exact same text?

Why are people willing to pay more for one product when another product can technically do more?

I’ve come to think the differentiator is the same one it has always been.

Taste.

Apple is probably one of the simplest examples. Apple is not always first, and sometimes it is quite late to the game: take the newly announced iPhone Duo. Apple is entering the foldable phone market in 2026, but the first commercially available modern foldable smartphone appeared back in 2018. Samsung launched the original Galaxy Fold in 2019 and has spent the years since iterating on the idea, reaching its eighth generation of foldable devices before Apple released its first.

There is nothing novel about the basic proposition of the iPhone Duo.

A phone that opens into a tablet-sized screen has existed for years.

And yet Apple’s entrance into the category immediately made the idea interesting again to people who had spent years largely ignoring foldables.

Since the passing of Steve Jobs, Apple’s argument has rarely just been innovation. More often, it is something closer to we think this is finally ready to be done our way.

With the Duo, Apple had years of other companies experimenting with hinges, aspect ratios, creases, multitasking, durability and the awkward relationship between a small outer screen and a much larger inner one.

Apple’s version arrives with software specifically redesigned around that transition, matching display proportions and the obsessive attention to animation and interaction that people already associate with its products.

Is it technically the best foldable phone?
No.
It is extraordinarily expensive, competing phones beat it in certain specifications and there are obvious first-generation compromises.

Why are people interested then?

Most consumers do not buy products by placing spec sheets next to each other and mathematically calculating a winner.

They buy based on how products make them feel.
They buy coherence.
They buy confidence.
They buy the feeling that somebody obsessed over details they themselves cannot even identify.
They buy taste.

And companies with enough of it can charge a premium for that feeling.

The obvious question, then, is where taste actually comes from.

I don’t think taste can be learned in the same way a technical skill can.
You can take a Python course and become better at Python. You can study networking and understand routing. You can follow a tutorial and learn how to deploy an application.

Taste is less linear. There is no ten-hour course that gives you good taste.

It has to be absorbed.

You develop it through years of consuming, noticing, and forming opinions on things. You discover what you like. Then, more importantly, you ask yourself why you like it. You realise that something you loved three years ago now embarrasses you. Your standards shift. You steal little ideas from hundreds of places and slowly combine them into something that feels like your own.

Films contribute to it.
Television contributes to it.
Books. Architecture. Clothes. Advertisements. Interfaces. Cars. Music. Hotels. Restaurants. Cities. Conversations. Cultures.

The more of the world you experience, the larger the library in your head becomes.

Eventually, you stop simply recognising things as good or bad.
You start to, whether consciously or otherwise, understand why.

That is taste.

And maybe some people are naturally more sensitive to it than others. But I don’t think anybody wakes up one morning with a completely formed point of view. Taste is the product of years of attention, followed by even more years of curation, rejection and iteration.

Which brings us to the obvious criticism:
Isn’t taste subjective?
Of course it is.
But subjective does not mean random.

Two people can disagree completely about whether a particular chair, website, outfit or car is beautiful while still recognising when something has been designed deliberately.

Good taste does not require everybody to like the same things.
It requires having a point of view strong enough that somebody does.

There is no universally perfect aesthetic, and there shouldn’t be.

The goal isn’t universal approval.
The goal is intentionality.
If you have taste, somebody out there will vouch for it.

And in a world where the mechanical act of making things becomes cheaper and easier every few weeks, the ability to decide what deserves to exist, what it should feel like, and what is worth obsessing over becomes more and more valuable.

AI can write, draw, code, design, research, analyse, animate, edit, compose, model, market, build, and ship.
When AI can do anything, taste becomes everything.