Chakrabarty notes that AI text signals skipped cognitive effort.
Part of it is also people protecting culture :) When I read something I want to know what the writer is thinking. So obviously when I smell AI writing it signals the writer has not spent time thinking about the topic. Why should I as a reader spend time reading it ?
Why is it so repulsive/cringe when someone uses obvious AI writing? I feel this, but can't decide if it's a feeling I want to defend. 1/
Users criticize obvious AI writing as cringe because it signals lack of author effort and violates reciprocity norms in communication, while some praise the analysis explaining these reactions.
Now, this need not be present in all domains of communication. But I think if readers feel like someone is violating a clear reciprocity norm, they'll likely want to punish the violator, by not reading, or even taking the time to denounce it.
Also means that as continual learning improves and AI-assisted writing becomes indistinguishable from the author's own writing (or at least requires too much mental computation for most of us), we will need a more robust sense of what is slop and what is kino.
I think this is just being Bayesian. People use AI in writing for two things mainly: 1) improving their writing after idea generation and 2) as a shortcut to idea generation (eg write me a few paragraphs making argument X and provide support). Most people are likely doing both 1) and 2).
When seeing writing, a person needs to decide whether the effort spent reading and thinking about is worth it in terms of learning new information from the writer. If the writer uses AI for 2), then there is little private signal or unique information—it is the same information the reader can generate themselves, probably more concisely, from using the AI model. So it’s perfectly reasonable to detect AI writing (either through detector or just feeling the cringe) and to stop reading.
The issue with using AI for 1) is that you are pooling with 2). It is very difficult to signal that you spent time and effort generating the ideas and have unique information to convey.
This is all to say this: it is perfectly reasonable to use AI as much as you want for writing, but it’s worth being aware that it is also perfectly reasonable for others to not read it.
@arindube Agreed, but I feel the same repulsion even when, say, reading a tutorial on something I want to learn.
@alexolegimas there's diminishing marginal utility to slop writing regardless of whether it was written with thought or not. good literature retains people - problem with AI editing is that retention falls off a cliff
@alexolegimas Another reason people use AI for writing is to save time and effort. For me that's the main use case. When I need something written and I don't care about the quality I can save time by delegating to AI. I rarely use it for ideas or for changing writing I actually care about.
@alexolegimas DFW actually addressed this quite some time ago
@alexolegimas The comp with coding is hard to resist. It's more acceptable to have an ai generated codebase; probably because it's easier to one shot writing than code. It may take a while, but ai writing that has been incrementally arrived at through iteration should become acceptable.
Yes, for sure: If you didn’t take the time to write it, why should I take the time to read it.
More importantly, imo: The byline is wrong! If I copied and pasted from Wikipedia but claimed my name in the byline, then everyone would say — correctly — that I plagiarized. I think people have a similar reaction to reading something that was copied and pasted from an LLM.
@alexolegimas yes. i don’t think the reader is being irrational here.
obvious ai writing collapses two very different cases into the same signal:
“i used a tool after thinking”
and
“i skipped the thinking”
from the outside, those can look identical.
so people leave.
@intellectronica Yes I meant for outward facing writing, that you want others to read.
I have dyslexia and rely on AI to QA what I type. As my focus is on a hotly debated field, consciousness, where language needs to be precise, having typos and incorrect words can be a real distraction from the point I'm trying to convey. But it is a double edged swords because I had a paper rejected because they passed it through an AI filter pre-screen which flagged it. Despite my saying I use it to help with my dyslexia it got no further.
@arindube I don’t think it’s about reciprocity or respect and more about importance. Not in how much time the sender put to you. Can it affect you, is it important, are there consequences from ignoring it or opportunities.
If I see “dear valued customer” I chuck letters, for example.
@alexolegimas I think this is a great point
@alexolegimas When using ai for idea generation the author does the selection and the prompt. It might still be valuable to publish the outcomes instead of the prompt alone.
@alexolegimas I've been calling it a parlor trick for years. But, it's an extremely useful parlor trick.
@alexolegimas As a writer, I couldn't agree more. I also think you just made what's roughly the best case against the theses you and @DKThomp laid out in the recent podcast ep. :)
@alexolegimas It also signals that the author isn't quite invested in their message. And my enthusiasm in reading something is usually upper-bounded by the conviction of the author in sharing it.
@alexolegimas
@alexolegimas Very good, agree with this. More precisely formulated version of my view than I had yet developed.
@MeganTStevenson @arindube That’s interesting. I don’t think I’d feel that way unless the human “author” claimed they’d written it themselves.
Which is closely related to my theory of why I dislike AI-written posts and papers: because the people passing it off as their writing are lying.
