AI building is the new doom scrolling.
Same dopamine. Same avoidance. Better disguise.
I’m three hours into a Sunday afternoon. Something functional sits on my screen. And I catch myself mid-keystroke with a question I should have asked before I started: Is this going to matter?
My brain has always moved faster than my hands could type. ADHD does that. Ideas fire constantly. The bottleneck was always execution. Building took time, which meant friction, which meant only the ideas I really cared about survived long enough to become real.
AI closed that gap.
Now my tools move at the speed of my brain. I thought this was the fix. The superpower I’d been waiting for. Every idea could become a prototype. Every itch could get scratched.
You probably see where this is going.
I have a Google Keep note titled “App Ideas.” It has over 20 entries. Most of them I could build in a weekend. Some of them I’ve already started. A few of them technically work.
None of them are going anywhere.
The thing about doom scrolling is you know you’re doing it. The guilt is built in. But building feels productive. You’re making something. You’re learning. You’re shipping.
Except when you step back, it’s the same pattern. Reach for the phone. Reach for the terminal. Quick dopamine hit. Avoid the harder question of what actually deserves your focus.
I traded one scroll for another. This one just has better optics.
The market doesn’t help. Everyone with an API key is shipping right now. Whatever idea I have probably already exists in some form. Three other people are building it this weekend, too. The log jam is real.
And the economics are genuinely weird. Why would someone pay for an app when they can prompt one into existence? I don’t know what the future of paid software looks like. Neither does anyone else.
So if building fast isn’t the advantage, what is?
I keep coming back to depth. Actually understanding a problem before solving it. Going narrow instead of wide. Picking one thing and getting genuinely good at it rather than having 20 half-finished repos.
The question I’ve started asking myself mid-spiral: how is this going to help me? When I look back on this time, will I feel like it was well spent?
Sometimes the answer is yes. Learning a new tool. Exploring a domain. Building something small that teaches you something big.
But often the answer is that I’m scratching an itch that doesn’t need scratching. And I could be doing something harder. Something that actually compounds.
The irony is that prioritization was supposed to get easier when execution got cheap. Instead, it got more important. When you can build anything, the skill is knowing what not to build.
I’m still figuring this out. The ADHD brain doesn’t shut off. The ideas keep coming. The tools keep getting better.
But I’m trying to catch myself more often. Pause before starting. Ask the question before I’m three hours into a doombuilding session.
Not every idea deserves a weekend. Some of them just need to stay in the Keep note.
Until next week,
Mike Watson @ Product Party
P.S. Want to connect? Send me a message on LinkedIn, Bluesky, Threads, or Instagram.
Brilliant framing. The paradox of cheap execution making prioritization harder is something I've felt but couldn't articulate. When I spun up my third side project in a week last month, it finally clicked that the dopamine loop was identical to scrollign Twitter. The disguise is perfectt because the output looks productive, but depth got replaced with breadth in a way that compounds into nothing.