I Know How to Build. I Just Don’t Know What to Build.
Knowing what AI can do has not made choosing what to build any easier. I may need less of an idea and more permission to play.
ai · building · creativity
It is the morning of my son’s second day at his new school. I am back at work. AI is everywhere and, for once, I feel on top of it.
I run workshops about what is possible with AI. I build agents that replace or augment parts of real work. Increasingly, it is not even engineering work. It is research, planning, communication, administration and all the small tasks that quietly fill our days.
There is a huge need for knowledge. People are adopting AI, but many still do not know what is possible or where to start. I help them see it and teach them how to do it.
The strange thing is that I feel a little lost myself. Not because I do not know what is possible. Perhaps because I know too much about what is possible. I want to build something useful for myself, but I keep helping other people build their things instead. Their problems are clear. There is a need, a context and someone waiting for the result. When I build for myself, everything is open.
I could build almost anything. So I build nothing.
Every idea has to be useful enough, interesting enough and important enough to justify the time. It quickly becomes a project, a product or a plan. The play disappears before I begin.
I already know how to discover what to build: by building. Start somewhere. Make something small. Fail. Learn. Change direction. Add another piece.
Like playing with LEGO bricks.
As a child, I did not sit down with a product strategy for my LEGO. I did not need to know what I was building before I started. I picked up a brick, connected it to another one and discovered the thing along the way. That kind of play is what I am missing. I know how to make things, but I seem to have forgotten how to make them without first knowing why they should exist. Knowing the answer—just sit down and play—has not made doing it any easier.
Perhaps the question is not what I should build. Perhaps it is how to stop demanding an answer before I begin. Maybe I do not need a better idea. Maybe I just need to open the box and pick up the first brick.
