Open any feed today, and you will see it: another app, another wrapper, another “I built this in a weekend with AI” post. Some of it is genuinely interesting. A lot of it is noise, things made not because they needed to exist, but because the person making them had access to a model and a free afternoon.
Experimenting is fine. Learning by building is how most of us got here. But somewhere between curiosity and ego, a strange culture has taken hold: build first, justify later. Ship something, anything, as long as it has “AI” in the description.
That is not building. That is performance.
The question that gets skipped
Before “how do I build this,” there used to be another question: should this exist? Who is it for? What problem does it actually solve? What happens to the people on the other side of it?
AI has not made those questions less important. It has made them more important because the cost of producing something has dropped so low that the friction which used to filter out bad ideas is gone. If it takes an afternoon to spin up a chatbot that scrapes someone’s writing, or a tool that automates a decision that affects real people, you owe it to those people to stop and ask whether you should.
Ethics is not a checkbox you add at launch. It is the thinking you do before you start.
Stop outsourcing the small stuff
There is a related habit worth naming. People are starting to ask AI to do things they could do themselves in five seconds to fix a typo in one word, rename a variable, add a comma. Every prompt is a small surrender of attention. Do enough of them, and you stop noticing your own work.
The point of these tools is to extend what you can do, not to replace the parts of thinking that keep you sharp. Reading your own code, sitting with a paragraph, fixing your own typo, these are not chores. They are how you stay connected to the thing you are making.
Use AI. But use it consciously.
I am not anti-AI. I use it daily, and it has changed how I work for the better. The argument here is simpler: use it with intent.
Before you start a new project, ask whether the world actually needs it. Before you delegate a task, ask whether you are gaining leverage or just losing the habit of doing it yourself. Before you ship, ask who could be harmed by what you built and whether you have thought hard enough about that to defend the answer.
Access is not a mandate. Capability is not a purpose. The most valuable thing you can bring to AI right now is not speed. It is judgment.
Build less. Think more. And when you do build, build something worth the energy it took to make it.
