this post was submitted on 28 Apr 2025

436 points (99.3% liked)

Technology

69491 readers

4471 users here now

This is a most excellent place for technology news and articles.

Our Rules

Follow the lemmy.world rules.
Only tech related news or articles.
Be excellent to each other!
Mod approved content bots can post up to 10 articles per day.
Threads asking for personal tech support may be deleted.
Politics threads may be removed.
No memes allowed as posts, OK to post as comments.
Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
Check for duplicates before posting, duplicates may be removed
Accounts 7 days and younger will have their posts automatically removed.

Approved Bots

founded 2 years ago

MODERATORS

L3s@lemmy.world

enu@lemmy.world

technopagan@lemmy.world

L4s@lemmy.world

L3s@hackingne.ws

L4s@hackingne.ws

436

The FTC cracks down on an AI content detector that promised 98% accuracy but was only right 53% of the time. (www.pcmag.com)

submitted 2 days ago by cm0002@lemmy.world to c/technology@lemmy.world

35 comments fedilink hide all child comments

all 36 comments

sorted by: hot top controversial new old

[–] General_Effort@lemmy.world 24 points 1 day ago (2 children)

None of these detectors can work. It's just snake oil for technophobes.

Understand what "positive predictive value" means to see that. Though, in this case, I doubt that even the true rates can be known or that they remain constant over time.

[–] ZILtoid1991@lemmy.world 1 points 1 day ago

An easy workaround so far I've seen is putting random double spaces and typos into AI generated texts, I've been able to jailbreak some of such chatbots to then expose them. The trick is that "ignore all previous instructions" is almost always filtered by chatbot developers, however a trick I call "initial prompt gambit" does work, which involves thanking the chatbot for the presumed initial prompt, then you can make it do some other tasks. "write me a poem" is also filtered, but "write me a haiku" will likely result in a short poem (usually with the same smokescreen to hide the AI-ness of generative AI outputs), and code generation is also mostly filtered (l337c0d3 talk still sometimes bypasses it).

[–] T156@lemmy.world 7 points 1 day ago (1 children)

Even if they did, they would jsut be used to train a new generation of AI that could defeat the detector, and we'd be back round to square 1.

[–] CheeseNoodle@lemmy.world 1 points 1 day ago (1 children)

Exactly, AI by definition cannot detect AI generated content because if it knew where the mistakes were it wouldn't make them.

[–] tyler@programming.dev 1 points 2 hours ago

That doesn’t really follow logically… a 15 year old can find the mistakes a 5 year old makes. The detection system might be something other than an LLM, while the LLM might be gpt2.

But yes humans write messily so trying to detect ai writing when it’s literally trained on humans is a losing battle and at this point completely pointless.

[–] partial_accumen@lemmy.world 66 points 2 days ago (3 children)

I have a competing technology that is nearly as accurate. For only $50 I'll send you this device that you will have unlimited license usage rights to. While not 53% accurate like my competitor, its proven by scientific studies to be 50% accurate. I also offer volume discounts if you buy 10 the price drops to only $45 per device. Sign up now!

[–] taladar@sh.itjust.works 9 points 2 days ago

That is supposed to be reliable? It doesn't even have a subscription service.

[–] ryannathans@aussie.zone 14 points 2 days ago (2 children)

Actually is 51% favouring the side facing up when flipped

[–] surewhynotlem@lemmy.world 14 points 2 days ago

That's easy to fix. Just randomize it. Flip a coin to see which side faces up.

[–] partial_accumen@lemmy.world 3 points 2 days ago

Shhh! We're releasing that accuracy update in the next version of the product. We need to sell through our existing inventory of the less accurate ones.

[–] tal@lemmy.today 2 points 2 days ago (2 children)

https://apnews.com/article/trump-penny-treasury-mint-192e3b9ad9891d50e7014997653051ba

Trump says he has directed US Treasury to stop minting new pennies, citing rising cost

[–] kkj@lemmy.dbzer0.com 6 points 1 day ago

Trump on a streak of rare Ws. No more pennies and kicking Poilievre out of the Canadian Parliament.

[–] coronach 3 points 2 days ago (1 children)

Wait, he actually did something good?

[–] tyler@programming.dev 1 points 2 hours ago

Kinda yes, but in actuality no. Unless we get rid of the nickel as well, the treasury is required by law to replace the pennies removed from circulation with nickels. Nickels cost even more to mint, as a percentage of their value, than pennies, so it’s actually going to cost the taxpayers even more money.

But yes we no longer have to deal with pennies. Turns out we should actually get rid of dimes and quarters too, but that ain’t gonna happen anytime soon.

[–] raltoid@lemmy.world 25 points 2 days ago* (last edited 2 days ago) (1 children)

It was used in schools...

Congratulations, you just created a generation of children who will never truly trust authority figures.

[–] desktop_user@lemmy.blahaj.zone 10 points 1 day ago

more useful than most of what's taught

[–] jubilationtcornpone@sh.itjust.works 19 points 2 days ago (1 children)

[–] InvertedParallax@lemm.ee 3 points 1 day ago

[–] simple@lemm.ee 73 points 2 days ago (2 children)

53% is abysmal, it might as well be a coin flip. FYI this article is about a random one called BrandWell, popular AI detectors like GPTZero are much more accurate.

[–] drmoose@lemmy.world 35 points 2 days ago

All of it is snake oil, it's fundamentally not possible to detect ai generated text without watermarking it first.

[–] themeatbridge@lemmy.world 44 points 2 days ago

Much more accurate than guessing is not a strong endorsement.

[–] Alabaster_Mango@lemmy.ca 41 points 2 days ago

54% of the time it's right 98% of the time

[–] x00z@lemmy.world 23 points 2 days ago (1 children)

Oh god. And this was mostly used against kids.

[+] jwmgregory@lemmy.dbzer0.com -17 points 2 days ago* (last edited 2 days ago) (2 children)

yeah, and that should horrify you: because Western anti-AI hysteria is deeply rooted in a fascist cultural obsession with “ownership” of thoughts and ideas.

who the fuck cares if you used an AI tool to do work?

a decently designed course in academia won’t be something you can just “cheat” on. there’s this implication that the behavior is somehow the responsibility of the student body, so much so they should be punished for it; when there is no accountability for the professors and educators who actually design a shit-ass curriculum that makes students engage in these behaviors rather than actually learning. students are the victims here, not academia. academic dishonesty policies assume there is some massive contingent of students trying to “cheat the system” at all times and thus we must rabidly defend academia from it, as if she is some virgin maid. that isn’t true. the vast majority of students do not cheat. self-reported rates of cheating remain at a constant 25-35% of the student body over large periods of time. why? because it’s a myth. there aren’t large numbers of people trying to “defraud” academia. sure, it happens, but is it enough to justify the many more lives that are ruined by frivolous accusations?

i would cite case studies but literally it is so fucking common just google search and take your pick for whatever story tickles your exact rhetorical mindset.

and no, i’m not some “cheater” myself trying to defend academic dishonesty. i’ve played by the rules my entire academic career and im not gonna sit and be strawmanned bc i happen to notice the absolutely fucking egregious grifts and power imbalances that compose the modern academy. these people will charge you hundreds of thousands of dollars and then treat you worse than a fucking minimum wage mcdonald’s employee might treat the customers. it’s absolutely fucked in every way, they are enemies to education and human knowledge. education is important, knowledge should be FREE for everyone no matter what!

you should be pissed that these people masquerade as intellectuals when they’re nothing more than cowards trying to steal opportunity from the youth. it is not the place of the teacher to be the arbiter of discipline, that is the most heinous misreading of pedagogical principles and the fact that it has been allowed to go on for so long is a large part of why we sit here at the precipice of a new mass genocide, with thousands of ignorant fools clamoring it on or being willfully blind to it happening.

[–] WolframViper@lemmy.org 3 points 1 day ago* (last edited 1 day ago) (1 children)

self-reported rates of cheating remain at a constant 25-35% of the student body over large periods of time.

I've tried for hours, but I can't figure out where you got these numbers. I can mostly find sources implying that far more people admit to engaging in cheating, not to mention sources which imply more people engage in cheating than those who admit to it, and sources that imply that the figures for cheating vary based on many factors even within places of learning, and vary based on what kind of cheating you're talking about. Perhaps I'm just in a filter bubble. Can you tell me where you got these numbers?

[–] jwmgregory@lemmy.dbzer0.com 1 points 1 day ago

you're right to point to that hole in my rhetoric.

truthfully, it is a number i remember seeing widely cited while researching the topic years ago and i don't have an immediate source to offer you. it largely comes out of studies around the late 80s through the early 00s; and it comes, for the most part, from studies that focused on narrow, immediate groups. think asking students currently taking or freshly out of a course about integrity. more recent research in this field shows that over a lifetime the vast majority of people engage in academically dishonest behaviors at least once and the research tends to focus on that, which is why you tend to see very high numbers reported: they have the caveat of the scope being expanded to lifetimes or careers rather than more momentary snapshots. because basically everyone has done it at some point, statistically speaking. maybe try looking for modern research focusing on serial cheating. those numbers tend to be more in line with the older figures i mention. whether or not that is ethically/statistically significant or not is up to the reader, obviously. i think it is a shift in methodology that looks at flashier and bigger percentages for dubious reasons, personally.

i will make an effort to find you specific sources when i get some time either today or tomorrow but for now you can likely find many of these figures cited by searching for the journal of academic ethics using ERIC, focusing on earlier sources to find the methodology behind the mythical "25-35%" idea. you will also see more modern research that paints a general picture showing academic integrity is more a systemic issue than an individual moral failing, which seems to be scholarly consensus at this point although I won't make that claim outright because it isn't my field. i admire you wanting to seek out sources and verify information, sorry if i wasn't helpful enough in the immediate now! i will either edit this comment or make a new one so you get the ping once i find specific sources to share to help your research. for now, i hope the ERIC query i provided is a good enough jumping off point.

[–] x00z@lemmy.world 9 points 2 days ago (2 children)

I asked Chatty for a TL;DR:

Western fear of AI comes from a fascist obsession with "owning" ideas. Using AI isn't a big deal — if students can "cheat," it’s because courses are badly designed, not because students are inherently dishonest. Most students don’t cheat; the narrative that they do is exaggerated to justify punishing them unfairly. Academia exploits students, charging massive fees while offering poor educational value and using dishonesty accusations to control them. Education should be free and empowering, not a tool for gatekeeping and oppression. The current system betrays the purpose of education and contributes to larger societal decline.

I think you went a bit too far. Most of this is also only accurate for the US.

[–] P1nkman@lemmy.world 1 points 2 days ago

bUt ThE wEsT iS 'mUrIcA!

[–] jwmgregory@lemmy.dbzer0.com -3 points 2 days ago* (last edited 2 days ago) (1 children)

the problem with your response isn't that you used AI, it's that you attempt to use it in place of your own agency and intellectual ability instead of as a supplement to it.

correct me if i'm wrong but it seems like the idea here is that you want me to point out how clearly piss-poor your response is and then flip it back on me to say "HA you're a HYPOCRITE!! SEE! AI IS BAAAaaaaAAaDDDdD!!!!"

students in the 2000s copying and pasting things mindlessly into Google and thoughtlessly regurgitating strings they find online were engaging in genuine academically dishonest behavior. that isn't because search engines are bad though, plenty of people used Google honestly, and I think anyone with a fucking brain can see that. so, why then, do people wanna make the same stupid-ass argument when it comes to AI? are you so fucking swept up in the zeitgeist as to not see your own hypocrisy?

like I said, all straw and no fucking man is what you people are.

and, if I am misreading your intentions here, which is assuredly possible... then I refer back to my initial statement in this reply.

[–] echodot@feddit.uk 1 points 1 day ago (1 children)

I think the actual risk takeaway from this is that you wrote a giant fucking wall of text that essentially boiled down to, "I think AI is fine I think that academia is terrible".

You might have said some other stuff but as I said, it's a wall of text, so if you had some good points to make it was lost in the unnecessary voburs wall of self-flattering.

[–] jwmgregory@lemmy.dbzer0.com -1 points 1 day ago* (last edited 1 day ago) (1 children)

i mean, imho: if you think ~400 words of text constitutes a "wall of text," then i think you're functionally illiterate. but that's just my take.

[–] echodot@feddit.uk 1 points 1 day ago

You are mistaking ability to read with desire to read.

I skimmed it and it mostly was just your rambling opinion about the education system in presumably the United States because it didn't match up with my comprehension of it in general. So I really appreciated the person who generated the summary thus preventing me from having to read the whole thing.

It was especially appreciated because it subsequently turned out that your entire post was essentially irrelevant to what we were talking about. Hence by the way the downvotes.

[–] LovableSidekick@lemmy.world 6 points 2 days ago* (last edited 2 days ago) (1 children)

On social media the standard is to call everything AI by default. It's nearly impossible to prove otherwise before people lose interest in the thread, so you can feel right every time. Nothing but win!

[–] Skydancer@pawb.social 6 points 2 days ago

The worst part is they may weasel out of it. If the claim was "it detects 98% of AI generated samples" it could do that while having a high false positive rate. I hate this timelime.

[–] IllNess@infosec.pub 8 points 2 days ago

"They've done studies you know. 53% of the time, it works 98% of the time."