Ten Thousand Dollars for the Domain. Zero for the Website.
A couple of months ago, a former client called and asked me to meet for coffee.
His daughter was opening a neighborhood bakery. Her specialty would be her grandmother’s dulce de leche cakes. A family recipe. Plenty of enthusiasm, and one of those stories that, over coffee in the right company, starts to sound like the beginning of a small empire.
He wanted to help. He asked if my agency could handle the bakery’s identity and digital presence.
He had a surprise, too. “I’ve already bought her the domain.”
Not a twelve-dollar domain you pick up while deciding whether to add hosting. This one was short. Memorable. Just right for the brand his daughter had in mind.
It had cost him ten thousand dollars.
I admit I shifted a little in my chair. A domain like that guarantees nothing. But it tells you something about the ambition behind a business.
A few weeks later, his daughter wrote. She’d been “playing around with artificial intelligence.”
She sent a link. A landing page. And it was beautiful. Really beautiful.
Warm colors. Elegant typography. Cakes arranged just so on a table. A kitchen filled with that golden light that seems to shine only through the windows of businesses that don’t exist yet. There was even a lovely grandmother watching over everything from the wall.
His daughter loved it. And it hadn’t cost her a thing.
The trouble is, a website can be beautiful and completely wrong.
A website is not a mirror. It isn’t there to show you a digitally flattering version of yourself. It’s there to have an effect on other people.
Artificial intelligence has made getting a result extraordinarily easy. That’s wonderful news.
The trouble starts when we mistake having something to show for having solved the problem underneath.
Anyone can now ask for a website for an artisan bakery, a coffee shop, a candle brand, or luxury lofts. Minutes later, they have something that, not long ago, took a team of designers, developers, writers, and photographers.
The risk isn’t that artificial intelligence can’t produce something different. It’s that ten thousand other people can ask for practically the same thing.
Soon the internet fills with different businesses dressed by the same tailor. The same compositions, gradients, smiles. The same statistical perfection.
To ask for something different, first know what difference you’re after.
“I love it” is not a strategy.
Do you want neighbors walking in tomorrow? Or are you building a destination for foodies, selling cakes to order, supplying restaurants, or creating a brand you can franchise?
You wouldn’t speak to a hipster hunting for their next Instagram discovery the way you’d speak to someone who tastes dulce de leche and remembers childhood in Argentina six thousand kilometers from home.
Every one of those decisions changes the website. Sometimes it changes the business, too.
The machine can’t know a business plan that doesn’t exist. It can’t optimize goals nobody has defined. It can’t know what return the person at the keyboard expects if they don’t know either.
And there’s something more dangerous. If you don’t understand what the machine is doing, you won’t recognize when it’s doing it wrong.
Imagine that tomorrow artificial intelligence starts writing flawless contracts. Would we stop needing lawyers? For many tasks, we’d probably need far fewer billable hours. That would be good news.
But there’s a difference between saying, “Write me a contract to sell cakes to restaurants,” and knowing what contract that business actually needs.
The same goes for accounting. The books can look immaculate and still be wrong. We’ll probably find out when a letter arrives from the tax authorities.
Communication can fail much the same way. But its mistakes rarely arrive by certified mail.
Nobody comes in. Nobody buys. They don’t remember the brand. The business just doesn’t grow.
And if we’ve handed over both the execution and the judgment, one question becomes much harder to face. Why?
Delegation is not abdication.
A business owner can delegate accounting and still understand the numbers. Rely on lawyers and still pay attention to legal risks. Hire designers, developers, and strategists without becoming one of them.
That’s why we have experts.
Delegating execution is reasonable. Delegating expertise can be efficient. Giving up judgment is something else.
For decades, we paid a premium for the ability to produce. To design, code, write, put together a campaign, edit a video, analyze large amounts of information.
All of that is getting cheaper at an extraordinary pace. And that’s worth celebrating.
But when something scarce becomes abundant value moves.
We’ll pay less for the ability to make things. More for knowing what’s worth making, who it’s for, what it’s meant to achieve, and what we should reject even when we "love it."
So perhaps the scene that matters isn’t a machine doing something badly. It’s a machine doing something flawlessly.
A beautiful screen. Perfect cakes. Golden light. That lovely grandmother watching over everything from the wall.
And someone who can look at it for a few seconds and say, “No, we’re not done yet.”
Because artificial intelligence can do more and more things for us.
Thinking about why we do them is still up to us.