> First write with lots of formatting.
> Then figure out how to remove it.
> Then put it back, if you want.
In the fifth grade, we were required to write an outline for our research project essay. Imagine my delight when writing the paper was as easy as copying the outline and adding a couple extra words.
That value of formatting into bulleted lists reminds me of the McPhee method of writing, which was shared last year on HN. He manipulates physical note cards to write, and I was manipulating digital ones.
You have a great site! I really enjoyed the articles I read - especially the one about your friend replacing you, that was absolutely fantastic and is one my favorite reads about AI thus far.
I hope your friend, having trained one of his personal AI to be you, at least consumes more dynomight content now!
I wish all us were as reflective as you are here:
"I keep finding myself unconsciously treating AI as an anomaly—as a weird thing that’s happening right now before the world goes back to being “normal”. But we aren’t going back. This is how it’s going to be. Like this but more so."
>I’m not really looking for something to read. I just want to skim an overview of the main theories. I’ve experimented with asking AIs to give the same information in various styles, and I reluctantly concede that the formatting helps. But that’s not reading.
[...]
>Namely, why do so many people today seem to write more like Exhibit A than Exhibit B?
Because they're writing for an audience that's not really looking for something to read. See Axios and Smart Brevity:
It is also notable that the Feynman lectures (3 volumes) write about all of physics in 1800 pages, using only 2 levels of hierarchical headings: chapters and A-level heads in the text. It also uses the methodology of sentences which then cumulate sequentially into paragraphs, rather than the grunts of bullet points. Undergraduate Caltech physics is very complicated material, but it didn’t require an elaborate hierarchy to organize.
I think about it a lot when reading markdown feature-driven writing or catching myself doing it.
I wind up using this kind of formatting while writing in Google Docs. It's easier to jump around with the nav bar if the structure is legible to the app. But it does make the document less readable.
This is unfortunate, because (and probably much to the article writer's chagrin), I am a fan of making use of the formatting tools available for a given platform. When convenient, I love using an em/en-dash. And, when I'm up for it, italicizing and bolding words to convey prosody. It just feels right.
Now, though, I may be dismissed as an LLM (probably as an older model that can't quite use them formatting tools effectively).
I've been called out more than once for using too much italics in my writing.
But the trick is I usually write like I would speak. This leads to italicizing any word or phrase I'd speak emphatically. (Which, yes, I've also been called out for doing a lot when I speak. So what; I've also been told I'm good at getting my point across. I'll take it!) In any text important enough to go through multiple revisions, or to be written from the start with multiple revisions in mind, this characteristic is diminished. But most text is more throwaway, just like most speech, so it gets left a little rough.
This also tends to feel pretty natural. If you read LLM-written text out loud, or the prose TFA is talking about, it... does not feel natural at all. So what I'm trying to say is: some level of emphasis is just fine. Don't overthink it.
I'm speaking more of excessive headings for small sections, which are themselves just bullet point lists, all of this highly nested. That's the main thing demonstrated in the post here.
The LLM writing is basically infodump. Lists dressed up in different ways (sections, bullet points, nested bullet points, and sequences of "[bold]Blah blah[/bold]: blah blah blah" sentences.
I think it's getting more and more common, quickly. The obvious inference is that people are using LLMs a lot and starting to mimic them, consciously or unconsciously. (Probably the latter: if people have weak internal models of how to write well, being around a lot of LLM text can probably influence them pretty quickly.)
Let me point out that in the first section you had radically different goals than in the second section. Try rewriting the “look how confusing this meaningless slop” is in prose and “look how clear this communication is” in outline and see what variable you’re actually illustrating.
• When writing for myself, I tend to open a buffer and just ramble. I write like I speak, which is how someone I know irl clocked my Hackernews account.
• My wife appreciates the writing I produce this way. Sometimes I just sit and read aloud to her.
• When presenting material for work, it tends to go across much more clearly when I write like a damn AI.
Inferences that can be drawn:
• Reading for pleasure is a thing
• Some of us prefer long-form when reading for pleasure
• However, when on the job we prefer parsimony and preset structure
Seriously hate how young people write these days. Like the OP covers, every sentence is on a new line, emojis everywhere, and too many memes. Does that make me old?
In the fifth grade, we were required to write an outline for our research project essay. Imagine my delight when writing the paper was as easy as copying the outline and adding a couple extra words.
That value of formatting into bulleted lists reminds me of the McPhee method of writing, which was shared last year on HN. He manipulates physical note cards to write, and I was manipulating digital ones.
I hope your friend, having trained one of his personal AI to be you, at least consumes more dynomight content now!
I wish all us were as reflective as you are here:
"I keep finding myself unconsciously treating AI as an anomaly—as a weird thing that’s happening right now before the world goes back to being “normal”. But we aren’t going back. This is how it’s going to be. Like this but more so."
[...]
>Namely, why do so many people today seem to write more like Exhibit A than Exhibit B?
Because they're writing for an audience that's not really looking for something to read. See Axios and Smart Brevity:
https://en.wikipedia.org/wiki/Axios_(website)#Content
https://www.axios.com/smart-brevity
Only knew it as the "Axios style" so far and was always weirded out by it. Their articles read more like a form letter than an actual text.
Good to know it has a name.
https://www.nngroup.com/articles/how-users-read-on-the-web/
It is also notable that the Feynman lectures (3 volumes) write about all of physics in 1800 pages, using only 2 levels of hierarchical headings: chapters and A-level heads in the text. It also uses the methodology of sentences which then cumulate sequentially into paragraphs, rather than the grunts of bullet points. Undergraduate Caltech physics is very complicated material, but it didn’t require an elaborate hierarchy to organize.
I think about it a lot when reading markdown feature-driven writing or catching myself doing it.
Now, though, I may be dismissed as an LLM (probably as an older model that can't quite use them formatting tools effectively).
But the trick is I usually write like I would speak. This leads to italicizing any word or phrase I'd speak emphatically. (Which, yes, I've also been called out for doing a lot when I speak. So what; I've also been told I'm good at getting my point across. I'll take it!) In any text important enough to go through multiple revisions, or to be written from the start with multiple revisions in mind, this characteristic is diminished. But most text is more throwaway, just like most speech, so it gets left a little rough.
This also tends to feel pretty natural. If you read LLM-written text out loud, or the prose TFA is talking about, it... does not feel natural at all. So what I'm trying to say is: some level of emphasis is just fine. Don't overthink it.
The LLM writing is basically infodump. Lists dressed up in different ways (sections, bullet points, nested bullet points, and sequences of "[bold]Blah blah[/bold]: blah blah blah" sentences.
When I've noticed this it's been in contexts where things lean against text being fully LLM-generated but... who the hell knows.
• When writing for myself, I tend to open a buffer and just ramble. I write like I speak, which is how someone I know irl clocked my Hackernews account.
• My wife appreciates the writing I produce this way. Sometimes I just sit and read aloud to her.
• When presenting material for work, it tends to go across much more clearly when I write like a damn AI.
Inferences that can be drawn:
• Reading for pleasure is a thing
• Some of us prefer long-form when reading for pleasure
• However, when on the job we prefer parsimony and preset structure