The AI Dopamine Trap: A Perfectionist's Best Friend and Worst Enemy

I want to be honest with you about something, because I think a lot of us are quietly living this and not saying it out loud.

The world feels dark right now. Not in a vague, general way — in a specific, you-can't-quite-look-away-from-it way. There are things happening at a systemic level that are genuinely unsettling, and if you're someone who pays attention, that weight doesn't disappear just because you close a tab. It travels with you.

And into that context comes AI — which, let's be honest, has its own unsettling edges. The nefarious potential is real. In the wrong hands, these tools are genuinely dangerous. I'm not going to pretend otherwise or skip past that to get to the good stuff.

But here's the thing I keep coming back to: the good stuff is also real.

"The creativity and productivity that AI unlocks — when you actually use it, not just theorise about it — is a genuine game-changer. And that matters, even when the world is hard."

What AI actually gives you back

I don't mean AI in the abstract. I mean the specific, practical experience of sitting down with a blank page that has defeated you for three weeks, typing a few sentences into Claude or ChatGPT, and suddenly having something to work with. That feeling of the block dissolving — of the idea that was stuck somewhere between your head and the page finally making it through — is not nothing.

For people who've spent years fighting their own perfectionism, who've lost entire mornings to the paralysis of "it's not right yet, I can't start yet," this is genuinely significant. AI removes the blank page. And for a lot of us, the blank page was the whole problem.

The productivity unlocks are real too. Tasks that used to take two hours take twenty minutes. Systems that required a team now run on one person with the right tools wired together. The business you were building in the margins of a full life starts to feel possible in a different way — not because AI does the thinking, but because it removes the scaffolding work that was eating your energy before the thinking even began.

But then there's the other thing

Here's what I didn't expect, and what I suspect I'm not alone in: AI responses are dopamine hits. Each one. Every time.

You put something in, something comes back. It's usually good. Sometimes it's surprisingly, delightfully good. And your brain — which is very good at wanting more of what felt good — immediately wants another one. So you ask a follow-up. And then another. And then you're refining the output that didn't actually need refining, just because the act of prompting and receiving is its own reward.

For a perfectionist, this is catastrophic in the most entertaining possible way. Because now you don't just have your own impossible standards to satisfy — you have an infinitely patient collaborator who will produce another version, any time, without complaint. One more draft. One more angle. One more variation. It never says "that's good enough, send it." It just keeps going, because you keep asking.

And for a procrastinator? Even better. You're not avoiding the task anymore. You're working on it. You're actively in dialogue with it. There are outputs. There is visible progress. It looks like productivity from the outside. It feels like productivity from the inside. And yet somehow an hour has passed and you're on iteration eleven of something that was finished at iteration two.

"You can never just get one hit. And therein lies the dilemma."

Why this is worth naming

I'm not saying this to warn you off AI. I'm saying it because I think it's more useful to understand what's actually happening in your brain when you use these tools than to just notice you're spending more time than you meant to and vaguely feeling bad about it.

The dopamine loop is a real mechanism. Unpredictable rewards — and AI outputs are variable enough to qualify — trigger dopamine release in a way that more predictable outcomes don't. Your brain is doing exactly what brains do. It's not a discipline failure. It's a design feature being activated by a very well-designed tool.

Knowing that doesn't make the loop disappear. But it does mean you can approach it more deliberately. You can decide in advance what "done" looks like before you start a session, so you have a finish line your brain isn't allowed to move. You can use AI for the parts where it genuinely replaces friction — and stay out of the loop when you're just refining something that was already good enough twenty minutes ago.

Holding both things at once

I think this is where a lot of conversations about AI go wrong — they ask you to pick a side. Either it's transformative and the naysayers are just scared of change, or it's dangerous and the evangelists are dangerously naive. Neither framing is honest.

The more useful thing — and the harder thing — is to hold the complexity. Yes, AI has genuine nefarious potential. Yes, it can be misused at scale in ways that are genuinely alarming. And also: for the woman sitting at her kitchen table at 7am trying to write the email she's been avoiding for two weeks, it can be the thing that finally gets it done. Both things are true simultaneously. The technology doesn't become safe to use just because you're using it well, and it doesn't become useless just because it can be misused.

What I've settled on, for myself: I use it where it removes friction without creating a different kind of friction. I notice when I'm in the loop versus when I'm genuinely working. And I try to be honest — with myself and anyone who'll listen — about both the genuine gifts and the genuine traps.

It's a game-changer. It's also a perfectionist's nightmare. Sometimes those are the same sentence.