Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Writing with LLM is not a shame

reflexions.florianernotte.be

Post reply on HN
Print view View original

121–130 of 155 posts

Re: Writing with LLM is not a shame

#121

Post by satisfice »

handoflixue wrote: post #109
satisfice wrote: post #85

The author of this piece commits a common mistake: analyzing AI use as if communication is nothing more than an isolated transaction. Instead communication is usually a process of creating and maintaining a relationship of some kind with other people. Here’s a thought experiment: Imagine if I handed you a $100 bill and asked you to examine it carefully. Is it real money? Perhaps you immediately suspect it is counterf…

Every day cashiers accept $100 bills on the basis that they pass the counterfeit tests, and every day society has failed to collapse from what you posit is a "hazard to the social contract"

Wow… that’s your counterargument?

First you are confusing the passing of a test with the actual non-occurrence of crime. My point is that passing a test for counterfeit money doesn’t make passing counterfeit bills legally or morally acceptable. Your point is what? That the tests on $100 bills always work, or when they don’t work that’s okay?

Surely you don’t mean that. I’m saying that I reject your AI-soaked fake work product even if I can’t see anything wrong with the work itself— because lying and stealing is not okay, but that’s what you are doing when you pass AI work off as your own thing.

Lots of criminals believe crime is no big deal. I am not trying to convince them, nor you, if bullshit is your creed.

Re: Writing with LLM is not a shame

#122

Post by johntash »

NicuCalcea wrote: post #91
CuriouslyC wrote: post #77

AI prose is mediocre right now. Too verbose, indirect constructions, passive, etc. That being said, it's actually a great editor and can pick out all those issues consistently. My workflow right now is to use AI for rough draft and developmental editing stages, then switch AI from changing files to leaving comments on files suggesting I change something. It is slower than letting it line/copyedit itself, but models d…

That's my main criticism as well. Even before we get to the ethical implications of AIs communicating on your behalf without a disclaimer, LLM writing is just poor and making me read through it is disrespectful of my time. I recently had a colleague send me a link to a ChatGPT conversation instead of responding to me. Another colleague organised a quiz where the answers were hallucinated by Grok. In some Facebook gro…

> I recently had a colleague send me a link to a ChatGPT conversation instead of responding to me.

I honestly would rather this than my colleague sending me text that is obviously from chatgpt, but not stating it upfront. Or even the "I asked chatgpt and it said this.." along with pasting 10 paragraphs of stuff they didn't even read to confirm it could be relevant.

Re: Writing with LLM is not a shame

#123

Post by viccis »

caconym_ wrote: post #112
viccis wrote: post #95

Earlier quoted context omitted.

Em dashes are fine in throwaway casual writing like internet comments or tweets or whatever. However, I think that, in any writing that is significant enough that LLM usage is scrutinized, they often just come across as a crutch to avoid more planned out sentence flow. I think it's actually a good thing that people are feeling like they should cut down on them.

Wrong. They show up in some of the best and most widely and intensely read prose that exists, with good reason. Of course, like any tool, they can be misused by people who don't know better.

You wanna cite any sources for that?

I don't know of any prose that relies on crutch dashes

Re: Writing with LLM is not a shame

#124

Post by kosolam »

nuancebydefault wrote: post #80
kosolam wrote: post #61

A relevant satirical post I stumbled on today much about the same subject: https://medium.com/@Justwritet/stop-competing-with-the-machi...

Interesting! How did you tell that it's satirical though?

Maybe it’s comedy and not satirical? I’m not an expert, it was a funny read so I figured it’s satirical..

Re: Writing with LLM is not a shame

#125

Post by ares623 »

redwall_hp wrote: post #64
tjpnz wrote: post #6

If you don't have time to write it I'm not going to make time to read it.

Exactly. LLM garbage is a straight up insult. 1. They deliberately chose to not take a few minutes to communicate with you, but expect something of you. 2. The hard part of writing is organizing thoughts into something coherent, not typing something out. If you don't understand something enough to write it in the first place, the LLM can't magically read your mind and understand what you want to say for you.

It’s even worse when you realize the equation is actually

1. “Author” spends 2 minutes prompting 2. 1000 people spend 15 minutes reading the slop. 15000 minutes wasted

Re: Writing with LLM is not a shame

#126

Post by KronisLV »

everdrive wrote: post #102

Writing with LLM is also not writing. In some abstract sense, it may be plagiarism. In another sense, you're robbing yourself of one of the most crucial parts of writing: improved cognition. Anyone who edits voice transcripts knows just how much a normal person wanders, pauses, misspeaks, etc when talking. The act of writing forces you to refine thoughts you would not have otherwise had.

I find it interesting that you could technically apply the same to writing code: you're producing software, but without engaging your cognition fully. Sometimes, when the complexity of the project is too high, it might be okay, to not let you get sidetracked and distracted. But if you do that regularly, I feel like that ability will atrophy.

Re: Writing with LLM is not a shame

#127

Post by acdha »

NicuCalcea wrote: post #91
CuriouslyC wrote: post #77

AI prose is mediocre right now. Too verbose, indirect constructions, passive, etc. That being said, it's actually a great editor and can pick out all those issues consistently. My workflow right now is to use AI for rough draft and developmental editing stages, then switch AI from changing files to leaving comments on files suggesting I change something. It is slower than letting it line/copyedit itself, but models d…

That's my main criticism as well. Even before we get to the ethical implications of AIs communicating on your behalf without a disclaimer, LLM writing is just poor and making me read through it is disrespectful of my time. I recently had a colleague send me a link to a ChatGPT conversation instead of responding to me. Another colleague organised a quiz where the answers were hallucinated by Grok. In some Facebook gro…

> I recently had a colleague send me a link to a ChatGPT conversation instead of responding to me.

I find this kind of thing interesting anywhere someone is being paid more than minimum wage: a really good way to make your boss think that they can replace you with ChatGPT is for you to perform it at ChatGPT’s level. I do give them points for not trying to hide it, but it really seems shortsighted not to consider that each time you do that, you’re raising the question of why they shouldn’t cut out the middleman.

Re: Writing with LLM is not a shame

#128

Post by Llamamoe »

sinuhe69 wrote: post #117
nicbou wrote: post #3

I think it's fair to use AI as an editor, to get feedback about how your ideas are packaged. It's also fair to use it as a clever dictionary, to find the right expressions, or to use correct grammar and spelling. (This post could really use a round of corrections.) But in the end, the message and the reasoning should be yours, and any facts that come from the LLM should be verified. Expecting people to read unverifie…

My natural reaction when I detect a writing of AI is now turning away. We have too much to read and too little time to waste on mimicry and not really what people thought or believed.

Yeah. Even when it appears and feels thoughtful, there's still no real intent underlying it, and frankly even if there was, the problem with AI output is that it ultimately gravitates towards similar outputs in a way that real human thoughts don't.

Re: Writing with LLM is not a shame

#129

Post by AlecSchueler »

viccis wrote: post #123
caconym_ wrote: post #112

Earlier quoted context omitted.

Wrong. They show up in some of the best and most widely and intensely read prose that exists, with good reason. Of course, like any tool, they can be misused by people who don't know better.

You wanna cite any sources for that? I don't know of any prose that relies on crutch dashes

Pretty much any good work of literature? Or any technical document? Framing them as "crutch dashes" doesn't instantly make them so.

Re: Writing with LLM is not a shame

#130

Post by Planktonne »

If you truly felt it wasn't shameful to write with AI, you wouldn't write an article trying to justify it that ignores a bunch of reasons people find it shameful.

No one is stopping you using AI, but there's an adolescent tone here of 'and you have to approve of me' that I don't care for. You have the right to act the way you wish; you don't have the right to be praised for it.

As it happens, I do think writing with AI (outside of translation as a method of last resort) is shameful. It's an embarrassing abdication of a basic task, a signal both that you can't complete basic tasks unaided, and that you don't understand that communication isn't solely performative.

If it's your right to use AI to write those super-difficult emails, then it's my right to judge you for it.

Post reply on HN