I've been noticing this pattern in myself for months and I finally snapped out of it.
You sit down with a simple idea. You ask ChatGPT about it. It gives you a really good answer. That answer mentions a few things you hadn't considered, so you ask about those too. Suddenly you're comparing architectures, databases, auth systems, deployment strategies... for an app that doesn't even exist yet.
Two hours go by and it genuinely feels like you made progress.
But you didn't build anything. You didn't even decide anything. You just explored endlessly.
And the weird part is ChatGPT (Or any AI) will never stop you. It won't say "you already researched this enough, just pick one." It won't tell you you're scope creeping or over-engineering your MVP. It'll happily keep answering forever.
So you end up drifting without realizing it.
Drifting - you came to build a task app, now you're designing an AI productivity ecosystem
Looping - you've compared React vs Vue vs Svelte three separate times and still haven't picked one
Second-guessing - you committed to Supabase an hour ago and now you're back to searching "Supabase vs Firebase" again
The dangerous part is that it all FEELS productive. You're learning things. You're thinking deeply. You're considering edge cases. But movement isn't direction.
AI solved the information problem. Now the problem is too much information and no guard rails. Nothing remembers what you already decided. Nothing tells you you're going in circles. It's just infinite helpfulness pulling you into deeper and deeper rabbit holes.
I've started asking myself one question whenever I'm deep in a ChatGPT session:
"What have I actually decided in the last 30 minutes?"
If the answer is "nothing," I'm probably not working.
I'm just spinning.
[link] [comments]