undefined

points

[-]

When "detectors" first started popping up all over the place, all of them rated the declaration of independence as 100% AI written, so... Yeah, these things just don't work. And what's even more dangerous is that people that don't understand how any of it works use these tools, and accuse people of using AI, sometimes with grave consequences. Students have been through this, at all levels of education.

by thesmtsolver213 hours ago|

parent|

[-]

I replied to a comment with a structured informal proof on HN.

A well established user here was adamant that I used AI because apparently humans never ever wrote proofs.

This was a while ago. Any well crafted human output is now being dismissively cast as AI if the reader is challenged by the output intellectually/politically.

by ycombinete8 hours ago|

parent|

[-]

It's the irl version of being accused of hacking in Counterstrike.

by roysting12 hours ago|

parent|

prev|

[-]

I’ve been accused of being a bot myself. I’ve started leaving some incorrect autocorrects in, hopping it would stave off people for whom it seems making accusations of being a bot is soothing or a coping mechanism when they’re confronted with information that does not fit into their world view or perception of self.

by jszymborski11 hours ago|

parent|

prev|

[-]

In fairness, the original crop of detectors were based on perplexity scores and were entirely useless (famously, the declaration of independence often came back as 100% AI generated).

I'm not convinced we'll ever have full-proof detectors, and certainly the false-positive rate will make them irresponsible for accusations of intellectual/academic fraud, I do think that LLMs are easy for folks to sniff out on average so I imagine it's possible to detect many instances.

Pangram's detector is anecdotally very accurate in my tests. This detector appears to be fine-tuned on a very small dataset (200 papers per subject), and suspect the problem might be in part that.

by yorwba14 hours ago|

parent|

prev|

[-]

They use a "threshold calibrated so pre-ChatGPT papers flag at 0.4%", so these things do work most of the time. It also means that there are known false positives, so for any given paper, scoring above the threshold isn't irrefutable proof of AI usage. But for things like estimating the overall proportion of AI writing, you only need to be correct on average, so individual false positives don't matter much.

by jasonfarnon13 hours ago|

parent|

[-]

It would be interesting to see how that .4% varies across fields on arXiv. I imagine some areas influenced LLM writing much more than others and are more susceptible to false positives. Anyway it's a little strange to me that these detectors do so poorly (at least by what I hear on this site). There is as much labeled data as you would need to train on pre-LLM human vs LLM text. If their out of sample errors are as good as has been claimed, the only methodological flaw I can think of is that post-LLM human text and LLM text have been converging.

by yorwba7 hours ago|

parent|

[-]

There's a table in the article breaking down the rate by field.

by sean_pedersen13 hours ago|

prev|

[-]

"Any attempt to build AI generated content (deep fake) detection systems is flawed, since the outputs of such a system may be used to train an even better fake data generator. This leads to an equilibrium state of digital uncertainty: nothing in the digital realm can be deemed as real anymore - only as digital. I do not care if a digital artifact is human or AI made - I only care if it is useful to me. Useful content is on point, factual and at best surprising (teaches something new)." - https://seanpedersen.github.io/posts/digital-uncertainty/

by miohtama7 hours ago|

prev|

[-]

It means machine text and human text cannot be distinguished from each other.

It's just text.

by dgellow17 hours ago|

prev|

[-]

Could it be that your papers are literally in the training set?

by cansofgrease18 hours ago|

prev|

[-]

It's been trained on a work and then distilled, the false positive noise has to be absurd.