Building has never been this easy.

I think that's exactly why so much of what gets built now is slightly wrong.

Two or three years ago, shipping anything was expensive. A feature meant weeks of engineering time, so at some point someone had to ask the uncomfortable question: do users actually need this? Not because teams back then were wiser or more user-centered. The cost simply forced the question. Friction was quietly doing the job of a product manager.

That friction is mostly gone now. You can describe something in the morning and have a working version by the afternoon. I don't think that's a bad thing. I build a lot myself and I would never go back. Being able to make five rough versions and put them in front of people in a week, instead of arguing about one for a month, is a real gift.

But the pull is real. When something can be built in a few hours, the default question quietly shifts. It stops being "what is the user struggling with?" and becomes "what can I make?" The two feel almost the same from the inside. Both produce output. Both feel like progress. Only one of them actually starts from a user.

It's a bit better when you are the user.

If you're building for yourself, you at least know where it hurts, so you tend to stop when the pain is gone. But most of us are not building for ourselves. And when you can't feel the cost, you keep adding.

That's what I see everywhere now. When adding something costs an afternoon, nobody fights it. One more toggle, one more AI button, one more panel. Each one is reasonable on its own. But the cost of adding didn't go to zero. It got transferred to the user, who now has to understand all of it. Removing a feature still takes a conversation. Adding one just takes a prompt. So products only ever grow in one direction.

The cost didn't disappear.

It moved. It used to live in engineering hours, where it was visible and easy to argue about. Now it lives in judgment. In deciding what not to build, and in understanding what someone is actually stuck on before you write the first prompt.

That kind of cost is invisible, which is probably why it's so easy to skip. Tools get faster every month. The question of whether something should exist at all doesn't.

So maybe the skill that matters now isn't building faster. It's being the person in the room who asks what this costs the user, and who's willing to say no to a feature that took only an afternoon. That used to be easy, because the afternoon used to be a month. Now someone has to choose to do it.