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

"No, I swear I wrote this."

revise.io

Post reply on HN
Print view View original

41–50 of 50 posts

Re: "No, I swear I wrote this."

#41

Post by potsandpans »

lelanthran wrote: post #15
sscaryterry wrote: post #11

A human should not have to or be compelled to prove to be human. The onus of proof here is the wrong way around, the premise fundamentally incorrect.

> A human should not have to or be compelled to prove to be human. The onus of proof here is the wrong way around, the premise fundamentally incorrect. Innocent until proven guilty is for a court. Outside of a court, people use all sorts of heuristics to determine authenticity and trustworthiness. Since so few humans ever wrote like the way LLMs default to, it's not unreasonable to refuse serious engagement with a pa…

[flagged]

Re: "No, I swear I wrote this."

#42

Post by Chu4eeno »

andai wrote: post #16

Didn't Pangram claim to have a >99% success rate? https://news.ycombinator.com/item?id=48667761 From their homepage: > Detect AI-generated content with 99.98% accuracy. Are they wrong? Though I ran the numbers, and even with a 0.02% false positive rate, that works out to about 6000 students falsely accused every semester, per university.

It's also confirmed in third-party studies like this https://arxiv.org/html/2501.15654v1 (which also confirms that humans that have experience with LLMs are very good at spotting them).

Re: "No, I swear I wrote this."

#43

Post by SwellJoe »

SwellJoe wrote: post #38
Ferret7446 wrote: post #35

Earlier quoted context omitted.

Do you know how many things throughout history people claimed confidently they could definitely tell the difference when it was later scientifically proven that they could not? And in all cases they believed that they were actually special and they were in fact correct despite the fact that this keeps happening.

I don't think I'm special. I think many people can tell the difference. Edit: Also, I'm surprised images have gotten to the point where I have a hard time detecting AI in some cases, and they got there more quickly than prose. I really thought prose would be the first to fall. Video is still detectable. Music still detectable (by someone that enjoys music and pays attention to it). But, AI prose still outs itself pre…

OK, this is a silly thing to do, but I wanted to be sure I wasn't imagining that I can tell when prose is written by AI. So I made a game. Turns out Claude Opus can fool me better than any other model, but even it can't fool me a majority of the time. I average about 85% accuracy on this (Claude prepared the corpus, I'm going in with very little foreknowledge). GLM is also very close to being able to convincingly write like a human. I'm not as good at detecting AI as I expected I would be, but I'm still pretty consistently able to detect AI prose.

https://prose-or-con.com

GPT 4o likes writing poetry with bees in it, for some reason. Qwen models are decidedly purple in their prose.

Re: "No, I swear I wrote this."

#44

Post by lelanthran »

potsandpans wrote: post #41
lelanthran wrote: post #15

Earlier quoted context omitted.

> A human should not have to or be compelled to prove to be human. The onus of proof here is the wrong way around, the premise fundamentally incorrect. Innocent until proven guilty is for a court. Outside of a court, people use all sorts of heuristics to determine authenticity and trustworthiness. Since so few humans ever wrote like the way LLMs default to, it's not unreasonable to refuse serious engagement with a pa…

[flagged]

Your post is unwarranted equivocation, a false dichotomy and disingenuous. Congratulations, you got a trifecta!

You're talking about a case where a mob is witch-hunting a person!

The actual case is a forum is rejecting a post.

You, and your form of rhetoric, are deliberately conflating the rejection of LLM messages and persecution of a person.

They are not at all the same thing, and I'd like to understand why you feel that LLM messages must be quietly allowed in specific forums.

After all, there are plenty of forums in which someone may post their LLM output, why is it so important that LLM output has to be allowed in some other forum?

All the objections I see boil down to one thing - "Those other forums allowing LLM output have very few human readers left. I want my LLM output to be viewed by humans, not by other LLMs".

Re: "No, I swear I wrote this."

#45

Post by SwellJoe »

ahofmann wrote: post #13
SwellJoe wrote: post #8

Trivial to simulate. Though I disagree that AI writing is impossible to differentiate from human prose, at least for now. It's still pretty obvious and still much worse than human prose (or, at least, much less interesting to read), though some are better than others, and I'm better able to spot writing from models I use regularly (Claude has a very distinctive style I can spot from a mile away, but that's true partl…

"It's still pretty obvious and still much worse than human prose" You have no idea how many false positives and how many false negatives you have in your judgement. It is indeed impossible to differentiate between badly written human text and somewhat good written llm text.

I wanted to be sure I wasn't lying to myself. So, I built this: https://prose-or-con.com/

I can still detect AI-written prose with pretty high confidence (I average about 85% accuracy after a bunch of rounds while testing, though I'm rushing because I'm testing), but it has gotten harder. Opus 4.8, GLM 5.2, GPT 5.5, and DeepSeek v4 Pro, are able to fool me sometimes. Mistral writes so weird it almost trips me up, but then I go, "oh, that's just Mistral". Qwen sometimes accidentally switches to Chinese (filtered out of the game, just a thing I found in testing). I'm doubling the corpus, currently, with a focus on the frontier models that have been able to fool me more often, so it'll probably get a little harder. GPT 4o is disgusting, just horrible saccharine nonsense (and, I guess that's the model that's caused the most psychosis with its sycophancy).

And, I have to concede that the best models prose has gotten better, especially when they're prompted to write in a specific style, like literary fiction or an encyclopedic article about a specific topic. Still flowery and verbose, though, and never really getting to the point (the AI doesn't want anything, and it shows).

Re: "No, I swear I wrote this."

#46

Post by andai »

lelanthran wrote: post #21
andai wrote: post #16

Didn't Pangram claim to have a >99% success rate? https://news.ycombinator.com/item?id=48667761 From their homepage: > Detect AI-generated content with 99.98% accuracy. Are they wrong? Though I ran the numbers, and even with a 0.02% false positive rate, that works out to about 6000 students falsely accused every semester, per university.

> > Detect AI-generated content with 99.98% accuracy. > Though I ran the numbers, and even with a 0.02% false positive rate They don't say that the false positive is 0.02%, only that the accuracy is 0.02%. All we know for certain is that the false positive and false negatives added together result in 0.02%.

Yeah, my bad. Well they don't exactly guard against that interpretation. "Up to 6000 falsely accused per semester!" doesn't exactly ease the mind!

Re: "No, I swear I wrote this."

#47

Post by andai »

mike_hock wrote: post #26
andai wrote: post #14

We're gonna have to bring back handwritten homework assignments. Oh, wait. https://images.ctfassets.net/kftzwdyauwt9/3bzFMXhknmq5TZvVL7... (From https://openai.com/index/introducing-chatgpt-images-2-0/ )

The irony is that a printout of a photo (well, "photo") of handwritten homework would be pretty easy to distinguish from genuine handwritten homework. (otoh it's also trivial to copy whatever ChatGPT wrote by hand without thinking about the assignment at all)

It would also be extremely educational. Handwriting is good for forming memory traces!

(Well, you'd be memorizing cringe and slop, but )

Re: "No, I swear I wrote this."

#48

Post by potsandpans »

lelanthran wrote: post #44
potsandpans wrote: post #41

Earlier quoted context omitted.

[flagged]

Your post is unwarranted equivocation, a false dichotomy and disingenuous. Congratulations, you got a trifecta! You're talking about a case where a mob is witch-hunting a person! The actual case is a forum is rejecting a post . You, and your form of rhetoric, are deliberately conflating the rejection of LLM messages and persecution of a person. They are not at all the same thing, and I'd like to understand why you fe…

Oh so you can't prove that you didn't murder someone, though many people are saying it.

Sorry, I don't debate things with criminals.

Re: "No, I swear I wrote this."

#49

Post by lelanthran »

potsandpans wrote: post #48
lelanthran wrote: post #44

Earlier quoted context omitted.

Your post is unwarranted equivocation, a false dichotomy and disingenuous. Congratulations, you got a trifecta! You're talking about a case where a mob is witch-hunting a person! The actual case is a forum is rejecting a post . You, and your form of rhetoric, are deliberately conflating the rejection of LLM messages and persecution of a person. They are not at all the same thing, and I'd like to understand why you fe…

Oh so you can't prove that you didn't murder someone, though many people are saying it. Sorry, I don't debate things with criminals.

> Oh so you can't prove that you didn't murder someone, though many people are saying it.

Your argument falls flat because no one, except you in your specific argument, and for the purposes of your argument, accused me of murder.

This example proves the opposite of what you wanted to prove: until you made up an accusation, you engaged with me. Your heuristics were sufficient, in your mind, to engage.

Your criteria is, presumably, "people I have baseless accusations against". Other people have different criteria, usualyl based on some evidence, not pure gut feel.

Re: "No, I swear I wrote this."

#50

Post by potsandpans »

lelanthran wrote: post #49
potsandpans wrote: post #48

Earlier quoted context omitted.

Oh so you can't prove that you didn't murder someone, though many people are saying it. Sorry, I don't debate things with criminals.

> Oh so you can't prove that you didn't murder someone, though many people are saying it. Your argument falls flat because no one, except you in your specific argument, and for the purposes of your argument, accused me of murder. This example proves the opposite of what you wanted to prove: until you made up an accusation, you engaged with me. Your heuristics were sufficient, in your mind, to engage. Your criteria is…

Please stop posting ai slop in response to whatever it is you think I'm arguing about
Post reply on HN