In my Oofday post, I shared a post in Hackernoon titled We Treated Potholes Like Software Bugs and Accidentally Built a Civic Hacking Playbook. The story is about a civic hack in Sofia.
Everything in the piece is excellent. The writing is vivid and clear. Its case is well-made. My only problem with it was suspecting it was written in some way by an AI. The tell:
- short sentences
- one-line paragraphs, often in lists
- contrasty (“not this, but that”) phrasing
- clever subheads (and/or clever everything)
- outline-like logical organization
So I had originality.ai examine it. The result: “We are 62% confident that the text scanned is AI-generated, NOT to be interpreted as 62% of the text produced is AI-generated.”
However, even if a piece is AI-generated, does it matter if it’s exactly what the author is trying to say?
Back in March, Daniel Barkhuff, MD, one of the most valuable sources of wisdom on the Web, copped to writing with AI. Specifically,
Let’s stop pretending. Everyone on Substack is using AI. If you think they aren’t, you’re high. I know because I am (using AI, not high), and I’ve been doing it for a while.
My process isn’t complicated. I sit down and write about a page, maybe 450 words. It’s not elegant. It’s not structured. It’s basically a brain dump. Half sentences, ideas that don’t quite connect yet, things I’d say out loud but that look ridiculous when you see them on the screen. It’s intellectual vomiting. A rough sketch of what I’m trying to say.
Then I paste it into AI and say something like, “Hey, give me an essay.”
And it does.
Three things about that:
- Not everyone on Substack (or anywhere lots of people write) uses AI. (I’m on Substack a bit, and I don’t write with AI.)
- This makes me think less of Dr. Barkhuff. Sorry, can’t help it.
- Knowing that he writes that way colors everything he’s written on his blog since I read that—and I read everything he writes, because his brain dump is the opposite of shit.*
Also bear in mind that the author of the Hackernoon piece, Bogomil Shopov – Бого, aka @bogomil, is writing to be read in thirteen different languages. Does it sound like AI in Bulgarian to begin with?
Hell, why not use AI to write something in your native language when you also know it will be translated into another twelve? (Is AI-style writing better for that? I’d bet it is.)
It should be clear by now that there is a learning loop that runs out through all of us, then through all the AIs that harvest and process our outputs, and then fed back to us in an idealized style that makes us (or at least me) reflect on ourselves and our own styles, while knowing that we too are being hacked by AIs and not just by other humans. Co-evolution at work.
By the way, I just ran this blog post through Originality.ai as a kind of controlled study. It said,
Likely Original
99% Confidence.
It also said I have only three scans left before I have to upgrade, which I probably won’t.
*I asked ChatGPT what “the opposite of shit” might be. I got, of course, a long-winded answer that boiled down to one word: gold. I can see the case, but it’s wrong, and I wouldn’t use it, except here, as a case in point.
Leave a Reply