It's a headline I read but didn't expect: ChatGPT usage is at an "all-time low" and a movement calling itself QuitGPT is said to have affected its popularity.
It made me curious, I didn't expect that headline in 2026. And yet — really, I shouldn't be surprised.
So what is QuitGPT?
QuitGPT is a reactionary movement against the widespread use of AI, specifically ChatGPT. It's not that AI is dangerous — it's that it replaces human thinking, human relationships, human creativity in a way that doesn't feel right.
Supporters are asking people to stop using ChatGPT — or at least to use it more thoughtfully. To question why they use it, really. To ask themselves what they lose when they forgo independent thinking.
It's an interesting pattern of backlash. We've seen it before — with smartphones, with social media. The technology explodes, "everyone" starts using it, then a group comes along and asks if we really want this.
Why I take it seriously
It would be easy to dismiss QuitGPT as Luddism — the fear of new technology dressed up in modern clothes. But I think it's more complicated than that.
There is a relevant discussion going on about what happens to our ability to think when we always have access to a tool that thinks faster and broader than we do. Brain muscles that are not used atrophy. Is it a problem that we stop struggling with formulations, stop pondering a difficult problem, stop enduring the unpleasant feeling of not knowing — because we can solve it in three seconds?
I don't know the answer. But the question is legitimate.
My relationship with AI and what it looks like right now
I have decided on a couple of principles that work for me.
I use AI to develop my thoughts, not to avoid thinking. That means I always formulate my own position first — or at least my own, sometimes, half-finished thoughts. Then I bring in AI to test them, challenge them, find holes in my reasoning. Not to generate them.
I always write my own lyrics. It's not a moralistic decision — it's a practical decision. My lyrics sound like me because they are me. If I were to delegate the writing, what makes them valuable would disappear.
I use AI openly. If I use it in a context where it's relevant, I say so. There's no reason to hide it.
TikTok's AI summarization fiasco
The same newsletter mentioned that TikTok’s AI creates ridiculously bad video summaries. One video was given the summary “Collection of various blueberries with different toppings.” It was probably a cooking video of some kind, but still.
It's fun, but it's also something to take seriously. We build systems to help us summarize, sort, and filter our information — and those systems are still pretty bad at context, irony, humor, humanity. They're good at the explicit and bad at the implicit.
The implicit is often the most important.
Why is this a Tech Thursday topic?
Technology is not just about what the tools can do. It's about what we choose to use them for, and what we choose not to use them for. It's an active choice, and we need to think through what we choose and also try to anticipate what effects different choices might have.
QuitGPT is probably not the answer. Blind adoption is definitely not the answer. Somewhere in between is a path that makes us sharper, not duller. More human, not more machine-like.
That's the path I try to take. Not always perfect, but fairly deliberate.
Have you noticed that your relationship with AI has changed in the past year — and if so, how?