A few thoughts on AI-assisted writing

Nabeel Qureshi recently tweeted that the winning entry of the 2026 Harper’s Bazaar short story competition was written by AI. He posted a screenshot from Pangram, one of those AI-writing detection platforms.

She sat on the bench opposite, the wood damp through her wool trousers – the Jigsaw ones from three years ago, back when she still believed dressing like a senior partner might hasten becoming one. Fake it till you make it, they said. But her Nani had warned her once, in that quiet voice that meant listen carefully: Watch what face you wear too long, beti. It starts to eat the one underneath. — From the essay.

A few things immediately came to mind.

First, I read the story in full. I enjoyed it. I thought it was well written, human or otherwise. Yeah, feel free to judge me.

Just before reading it, I had read a short note by Venkatesh Rao. The point he was making in the note was that it doesn’t really matter if someone use AI to write, what matters is the quality of the ideas and if they truly “delved” in their ideas and then day something.

I agree.

We are living through a strange twilight zone where many of the old frames we used to make sense of the world are slowly breaking down. AI is now part of everyday life, and a lot of the reactions to it seem familiar.

Human beings have an instinctive aversion to change and new things. We crave homeostasis—not just biologically, but psychologically. Whenever something new arrives and threatens to remake the old certainties we use to navigate the world, there is resistance.

Right now, AI is that thing.

I’m sure I’m in the minority here, but I like thoughtful AI writing.

To me, it doesn’t particularly matter whether a human writes something or an AI writes something, provided there is a human somewhere in the loop steering the process. If a person is using AI as a tool to express an idea, communicate an experience, clarify a thought, or share some hard-earned wisdom, I don’t really see the problem.

In fact, I suspect AI may trigger a small renaissance of expression.

There are countless people with fascinating experiences, original insights, unusual expertise, and genuine wisdom who simply do not have the writing skills to talk about those things. AI lowers that barrier. It allows people who otherwise would have remained silent to write. That’s a bloody good thing!

And if the final result gets flagged as “100% AI-generated” by some detector, I honestly don’t care.

In fact, after recording this note, I took a transcription of my own voice note, put it through Pangram, and, rather amusingly, it came back as 100% AI-generated. Even though I had written it myself.

The irony writes itself.

I care about whether the underlying idea is interesting.

I care about whether the author has something worth saying.

I care about whether the piece taught me something, challenged me, amused me, or helped me see the world differently.

The only exception to all of this is effortless slop.

By slop, I don’t mean AI-assisted writing. I mean textual garbage. Alphabetic diarrhoeal discharge produced without intention, curiosity, effort, taste, or heart. Content generated simply because someone wanted to fill a page, feed an algorithm, or manufacture engagement.

That is slop.

And the purveyors of such gunk deserve every bit of criticism they receive.

What I find amusing is that “this was written by AI” has increasingly become a kind of intellectual dunk. A rhetorical shortcut. It’s a way of dismissing something without engaging with it and that’s a travesty.

There’s an old refrain that goes something like this:

“If you didn’t take the time to write it yourself, why should I take the time to read it?”

I understand the sentiment. But I don’t entirely buy it.

Writing is thinking. Or at least, that’s the part I care about.

As long as somebody has wrestled with an idea, explored it from multiple angles, refined it, challenged it, and used AI as a partner to help translate those thoughts into words, I see no reason to dismiss the result.

And the more I think about it, AI-assisted writing doesn’t seem very different from AI-assisted coding.

Does it really matter if an experienced developer types every line of code himself, or if he oversees a bunch of AI agents and all you see at the end is a piece of software that works exactly as advertised?

Would you stop using the software because AI generated most of the code?

Probably not.

You care about whether the software works.

Likewise, when I read something, I care about whether it contains an interesting idea, a useful insight, a memorable observation, or a perspective I hadn’t considered before.

The real question isn’t whether AI wrote it.

The real question is whether there was intentionality.The real question is whether there was intentionality.

Whether there was something worth saying in the first place. Whether someone did the mental taekwondo and jujutsu with their thoughts and then knead those thoughts into words using AI. It’s just like making bread

Everything else feels secondary.

But to each their own.

Agree with what’s Alex Tabarrok is saying:

We should treat this just as if another Erdos problem has been solved by an AI. Another literary prize has been won by an AI! Amazing!