Posted on 01/13/2023 9:29:01 AM PST by BenLurkin
The app works by looking at two variables in a text - perplexity and burstiness - and it assigns each of those variables a score.
First, the app measures how familiar it is with the text presented given what it has seen during training. The less familiar it is then the higher the text's perplexity is, meaning "it's more likely to be human-written", Mr Tian said.
It then measures burstiness by scanning the text to see how variable it is. For example, does the text have a mix of short versus long sentences? Or does the writing appear to be more levelled and uniform?
"If you plot precisely over time, a human-written article will vary a lot," Mr Tian said. "It would go up and down, it would have sudden spikes."
He is still working on improving GPTZero, but he has released a beta version for public use. In a tweet, he demonstrated how the app can successfully sniff out the difference between an essay published in the New Yorker magazine versus a letter written by ChatGPT.
He said he has also since tested it out by feeding the app BBC articles written by journalists, versus articles written by ChatGPT using the same headline as a prompt. (Mr Tian formerly worked with the BBC's investigations unit). He said the app successfully guessed the difference between the texts with a less than 2% false positive rate.
Since its launch, Mr Tian's app has been used by thousands of people. He said he has since been contacted by teachers and university admissions officers from around the world who are interested in how it works.
(Excerpt) Read more at bbc.com ...
Holden : The tortoise lays on its back, its belly baking in the hot sun, beating its legs trying to turn itself over, but it can’t. Not without your help. But you’re not helping.
Leon : [angry at the suggestion] What do you mean, I’m not helping?
Holden : I mean: you’re not helping! Why is that?
Nice.
ChatGPT is actually more amazing than even this article depicts. This video describes how a college professor in Cosmology gave his final exam word problems to ChatGPT and it did as well as an average (but not top) student. If students can use materials on open book exams with calculators, when will they start using ChatGPT to answer word problems?
Let me tell you about my mother....
There’s a simpler solution. Go old school. Short form essays must be completed in class, written longhand on paper in one 40 minute class period. Research papers must be fully sourced with footnotes, using only pre-approved sources available from the school library.
Anybody trying to use an AI with those rules just fails anyway.
Blade Runner!!!!!
Will the app also be able to test if I’m a lesbian?
Great! Using one algorithm to detect the workings of another algorithm. Kids, you’ll have to write your own essays now!
The Voight-Kampff test.
Bladerunner has long been my all time favorite movie. I have owned the sequel for many years but have never watched it, for fear it may forever foul the original.
This movie was so ahead of its time.
To test this... I copied the last few months of your posting history and let the app analyze it. I think that the results might have been skewed by your canned hamster gif.
It concluded that you were a bipolar transexual with hamster hating tendencies. Let us know if that is accurate.
“I’ll tell you about my mother...”
>Holden : The tortoise lays on its back, its belly baking in the hot sun, beating its legs trying to turn itself over, but it can’t. Not without your help. But you’re not helping.
>Leon : [angry at the suggestion] What do you mean, I’m not helping?
>Holden : I mean: you’re not helping! Why is that?
You can do the same test on people on the margin between 90 and 80 IQ. Just ask things like “How would you have felt if you hadn’t eaten breakfast yesterday?” Higher than 90s tend to immediately say, “Hungry.” Lower than 90s tend to tell you about whether or not they ate yesterday.
Kids these days. Why can’t they just be honest like us and pay another human being to write their essays for them?
Opposing AI when it is used to compose essays is one day going to be viewed the same as opposing:
*Using a typewriter to write an essay instead of writing it by hand.
*Using a computer to write an essay and using the computer to check grammar and spelling before finishing it.
*Using the internet instead of a library for research.
*Using You Tube for research.
Etc.
The better practice is to identify ways to use AI instead of embracing one’s inner Luddite by opposing it.
They might as well try to develop an algorithm to see if the students used calculators.
ChatGPT is now a tool available to them. The world needs to adapt and put it to good use.
Great movie and scene.
I remember having to write some papers in class, where you could bring in only something to write with.
“The better practice is to identify ways to use AI instead of embracing one’s inner Luddite by opposing it.”
Chatgpt is revolutionary. It can be used for programming, web design, business reports (PowerBI scripts), photography videos, songs, poems, articles, etc...
Combine this with MidJourney and art has changed.
You’re right, this is a bandwagon that every techie needs to jump on to stay ahead.
Now as to defeating these AI Sniffers all someone has to do is run another AI Program called Jasper to rewrite the text. It renders these sniffers useless.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.