It Is Getting Easier to Spin Up Ideas Than Finish Them

The ideation pace has gotten out of control.

I have a habit of thinking through side project ideas outside of work. Apps I want to build. Tools I have been meaning to make. Usually that early framing work is slow. You sit with the idea. Figure out what it actually is. Identify whether it is worth building.

ChatGPT has made that part fast.

Name it, frame it, sketch the rough product thinking. What used to take a full afternoon now happens in an hour of focused back-and-forth. The early framing layer is just cheaper.

The execution has not gotten cheaper.

That is the gap that has opened up. Ideation cost is down. Shipping cost is UNCHANGED. So I have more ideas in a more developed early state than I can actually do anything with. The "things I could build" backlog is longer. The shipping rate has not moved.

That is not a complaint about ChatGPT. It is just what this phase looks like. The part of the work where AI helps most right now is early and upstream. Figuring out what you are building. Not building it.

The actual code still lives somewhere else. Context is still managed manually. The gap between a well-framed idea and something that ships is exactly as wide as it was.

This is where the leverage is right now: earlier thinking, faster. The execution gap is still real and I have no reason to think it closes on its own.

But the ideas keep coming.