It Wasn't AI

#1

Post by dbrereton » Mon, May 15, 2023, 5:07 PM UTC

Re: It Wasn't AI

#2

Post by tikkun » Mon, May 15, 2023, 5:13 PM UTC

This is a great idea. Huge problem. Most of these AI detectors have poor accuracy, but it seems like education institutions treat them as truth.

Re: It Wasn't AI

#3

Post by minimaxir » Mon, May 15, 2023, 5:31 PM UTC

Context for the website creation: https://www.reddit.com/r/ChatGPT/comments/13hi5y6/i_fed_gptz...

Re: It Wasn't AI

#4

Post by krossitalk » Mon, May 15, 2023, 5:33 PM UTC

Ironic that I had to pass Cloudflare "Are you human" verification.

Re: It Wasn't AI

#5

Post by simonw » Mon, May 15, 2023, 5:38 PM UTC

On the one hand, this is sorely needed: AI detection software will inevitably be mostly snake oil.

Academia and education desperately wants this software to work! As a result, selling them something that doesn't work is going to be very profitable.

The most obvious problem with this class of software is how easy it would be to defeat if the students could access it themselves: generate some text, run it through the detector, then fiddle with it (by manually tweaking it or by prompting the AI to "reword this to be less perfect") until it passes.

Which means these tools need to not be openly available... which makes them much harder to honestly test and evaluate, making it even easier to sell something that doesn't actually work.

But... I don't think this site is particularly convincing right now. It has spelling mistakes (which at least help demonstrate AI probably didn't write it) and the key "How AI Detection Software Works" page has a "Coming Soon" notice.

The "examples" page is pretty unconvincing right now too - and that's the page I expect to get the most attention: https://itwasntai.com/examples

It looks to me like this is still very much under development, and is not yet ready for wider distribution.

Re: It Wasn't AI

#6

Post by morkalork » Mon, May 15, 2023, 5:40 PM UTC

tikkun wrote: post #2

This is a great idea. Huge problem. Most of these AI detectors have poor accuracy, but it seems like education institutions treat them as truth.

How many essays are written and graded every year? Even with a 0.0001% FPR, how many students would be facing the severe punishments for plagiarism like auto-failing a course or even expulsion? Ironically enough, using AI to make such decisions seems like one of those unethical use cases.

Re: It Wasn't AI

#7

Post by jvanderbot » Mon, May 15, 2023, 5:40 PM UTC

Any sufficiently good AI detector will soon become part of the training process.

Re: It Wasn't AI

#8

Post by yieldcrv » Mon, May 15, 2023, 5:42 PM UTC

oooh I like that, the student can sue for copyright infringement because the teacher uploaded their work and proved that they uploaded it?

sounds like a simple sublicensing clause imposed on the student will fix that, but the next few semesters a few examples can be made of the teachers and institutions

will pay off that tuition

Re: It Wasn't AI

#9

Post by moffkalast » Mon, May 15, 2023, 5:45 PM UTC

Has anyone tried using an LLM as an LLM detector yet?

Re: It Wasn't AI

#10

Post by inconceivable » Mon, May 15, 2023, 5:45 PM UTC

this entire line of reasoning (using AI to detect AI, with disastrous results) is ripe for a giant lawsuit. a sufficiently wealthy school is bound to accuse a sufficiently wealthy student at some point.