I've been on a crusade against sloppy use of AI recently. A conversation between my brother and I spawned the idea of "emoji reaction dog whistles" and I couldn't stop thinking about the idea until I wrote it down. Not meant to be taken very seriously, but not entirely UNseriously either
> Selfish LLM usage: A situation in which someone uses an LLM to save their own time at the expense of other people’s time, resulting in a net productivity loss.
Brilliant. I would like to expand this definition slightly in order to include other problematic LLM usage.
> Selfish LLM usage: A situation in which someone uses an LLM to generate text or media which elicits or demands other people's emotional, cognitive or temporal input with the aim of saving similar resources for themselves, resulting in an unfair social exchange.
I think the better solution would be to treat every answer as if it were written by a human and then hold the person accountable for any mistakes. I’ve often seen in academic contexts that people have published texts with completely fictitious citations. In such cases one should confront the person as if they had done it on purpose. No one can accidentally invent entire books and authors. The same goes for software projects, where some people submit code that clearly comes from an LLM and hasn't been properly reviewed beforehand. If there are security vulnerabilities in that code, you can accuse the person of having done it on purpose. Anyone who submits AI code without labeling it as such is affirming that they understand the code and have thoroughly reviewed it.
I get pretty aggressive about this. I tell them if they expect human effort in review, then they should put some human effort into what they send me. Then I defuse it by ending on a positive tone with some tips about how to prompt better.
It’s worked well so far. Most people only send me that slop once.
The irony of someone writing a blog post purely to rage-bait (or rage-engage?) against "slop grenades." Slop grenades can also be made by humans you know, and I'd argue this post is at least a flashbang.
FTA: "Opinions vary on the usefulness of LLMs, but hopefully we can all agree that using them selfishly is cringe."
LLM = Bad, Human = Good
LLM's output can be useful and informed. If your going to deny that then I dont know what to tell you to convince you otherwise.
How about we get back to holding humans accountable, for the LLM's output and their own. I can point you at people who are using LLM's to make actual art, and it is GOOD (strong stories, social commentary and so on). I can point you to lawyers who are using LLM's to great success. Reading and redlining documents is part of being good at the job and trained into them by trade.
A lot of obvious LLM speech is obvious, but please spare us academics and former academics who learned how to use an em- and en-dash properly. Proper writing in general is being derided by younger folks who are too uneducated to tell the difference. I recently had someone claim my company's website, which is virtually unchanged since 2020, was "obviously AI." Yeah, kid, except you can go on Internet Archive and see it before ChatGPT was dropped.
But it doesn't matter. It's just playing chess with pigeons at that point.
I share your sentiment of not being overly judgmental when you see an en/em dash but this feels overly reductive and hostile:
> Proper writing in general is being derided by younger folks who are too uneducated to tell the difference.
There are legitimate differences in writing styles between age groups and cultures and being overly prescriptive in what is “right” vs “wrong” doesn’t help your argument IMO
I think you are misconstruing what I said. I'm not talking about other cultures, I'm talking about the youth who turn in essays that mimic their texting style. I should know: I spent years grading their essays and helping them fashion a more professional writing style.
It's true that there are differences but I think it's simultaneously the case that the practice in any form is on the decline at the population level. No idea what the expected endpoint might be though.
This type of mentality is really not helping anyone. If you really think shaming people works, I don't know what planet youve been on for the past 10 years. And I reallllly dont trust any of you to accurately determine intent lol.
The people going around pointing fingers about AI often get it wrong. Just listened to a podcast in which the host baselessly accused a film studio of using AI in their marketing and they later had to apologize because GUESS WHAT? An actual artist with feelings made the poster she was talking about. I'd honestly rather let a hundred AI fraudsters off the hook than falsely accuse one person of using it when they actually made a genuine effort.
I just started referring to those coworkers as Claude directly. Not like, at the start of the message, but after the first line? Nobody’s noticed yet.
I've been on a crusade against sloppy use of AI recently. A conversation between my brother and I spawned the idea of "emoji reaction dog whistles" and I couldn't stop thinking about the idea until I wrote it down. Not meant to be taken very seriously, but not entirely UNseriously either
I’m a particular fan of the ai-for-brains emoji in Slack
https://slackmojis.com/emojis/63197-ai-for-brains
> Selfish LLM usage: A situation in which someone uses an LLM to save their own time at the expense of other people’s time, resulting in a net productivity loss.
Brilliant. I would like to expand this definition slightly in order to include other problematic LLM usage.
> Selfish LLM usage: A situation in which someone uses an LLM to generate text or media which elicits or demands other people's emotional, cognitive or temporal input with the aim of saving similar resources for themselves, resulting in an unfair social exchange.
Just reply “I gave this to Claude and it didn’t understand what you meant - can you rephrase?”
I think the better solution would be to treat every answer as if it were written by a human and then hold the person accountable for any mistakes. I’ve often seen in academic contexts that people have published texts with completely fictitious citations. In such cases one should confront the person as if they had done it on purpose. No one can accidentally invent entire books and authors. The same goes for software projects, where some people submit code that clearly comes from an LLM and hasn't been properly reviewed beforehand. If there are security vulnerabilities in that code, you can accuse the person of having done it on purpose. Anyone who submits AI code without labeling it as such is affirming that they understand the code and have thoroughly reviewed it.
I get pretty aggressive about this. I tell them if they expect human effort in review, then they should put some human effort into what they send me. Then I defuse it by ending on a positive tone with some tips about how to prompt better.
It’s worked well so far. Most people only send me that slop once.
This is neat. I'd love to figure out a sequence of emojis that triggers the LLM in ways that puzzles a human.
thank you, I'll put this PR checklist in my claude.md
The irony of someone writing a blog post purely to rage-bait (or rage-engage?) against "slop grenades." Slop grenades can also be made by humans you know, and I'd argue this post is at least a flashbang.
Well...
FTA: "Opinions vary on the usefulness of LLMs, but hopefully we can all agree that using them selfishly is cringe."
LLM = Bad, Human = Good
LLM's output can be useful and informed. If your going to deny that then I dont know what to tell you to convince you otherwise.
How about we get back to holding humans accountable, for the LLM's output and their own. I can point you at people who are using LLM's to make actual art, and it is GOOD (strong stories, social commentary and so on). I can point you to lawyers who are using LLM's to great success. Reading and redlining documents is part of being good at the job and trained into them by trade.
Stop blaming the tool for peoples poor behavior.
>> using them selfishly is cringe.
> Stop blaming the tool for peoples poor behavior.
Are you then in agreement with the author?
[flagged]
A lot of obvious LLM speech is obvious, but please spare us academics and former academics who learned how to use an em- and en-dash properly. Proper writing in general is being derided by younger folks who are too uneducated to tell the difference. I recently had someone claim my company's website, which is virtually unchanged since 2020, was "obviously AI." Yeah, kid, except you can go on Internet Archive and see it before ChatGPT was dropped.
But it doesn't matter. It's just playing chess with pigeons at that point.
I share your sentiment of not being overly judgmental when you see an en/em dash but this feels overly reductive and hostile:
> Proper writing in general is being derided by younger folks who are too uneducated to tell the difference.
There are legitimate differences in writing styles between age groups and cultures and being overly prescriptive in what is “right” vs “wrong” doesn’t help your argument IMO
I think you are misconstruing what I said. I'm not talking about other cultures, I'm talking about the youth who turn in essays that mimic their texting style. I should know: I spent years grading their essays and helping them fashion a more professional writing style.
It's true that there are differences but I think it's simultaneously the case that the practice in any form is on the decline at the population level. No idea what the expected endpoint might be though.
This type of mentality is really not helping anyone. If you really think shaming people works, I don't know what planet youve been on for the past 10 years. And I reallllly dont trust any of you to accurately determine intent lol.
The people going around pointing fingers about AI often get it wrong. Just listened to a podcast in which the host baselessly accused a film studio of using AI in their marketing and they later had to apologize because GUESS WHAT? An actual artist with feelings made the poster she was talking about. I'd honestly rather let a hundred AI fraudsters off the hook than falsely accuse one person of using it when they actually made a genuine effort.
> This type of mentality is really not helping anyone. If you really think shaming people works,
Scroll down past the emojis to the actual interesting part of the article with positive reinforcement, empathy, etc.
Yeah i read it. As I said, I dont think any of you can accurately determine intent so its a moot point.