ChatGPT detector could help spot cheaters using AI to write essays

  • 📰 newscientist
  • ⏱ Reading Time:
  • 25 sec. here
  • 2 min. at publisher
  • 📊 Quality Score:
  • News: 13%
  • Publisher: 51%

Technology Technology Headlines News

Technology Technology Latest News,Technology Technology Headlines

A tool called GPTZero can identify whether text was produced by a chatbot, which could help teachers tell if students are getting AI to help with their homework

A tool built using the AI behind ChatGPT can help extract information from scientific paper abstracts.

It could help researchers identify important information across thousands of articlesThe European Union is set to create the world's first broad standards for regulating or banning certain uses of artificial intelligence in 2023University students tend to get lower grades if their classes and lectures begin early in the morning, because they struggle to wake up early enough to attend themAlgorithms can pore over astrophysical data to identify underlying equations.

 

Thank you for your comment. Your comment will be published after being reviewed.
Please try again later.

GPTZero's effectiveness reduces when you ask ChatGPT to rewrite something. I am writing notes for blue plaque tours this year, for free as a series of meetups. I ask ChatGPT for ten bullet points on a person/subject, ask to rewrite and hey presto - human like text!

What if students were required to write essays with pen and paper, like exams, in the presence of the professor/teacher/aides. Maybe that’s one place where even basic computers are unneeded tech. I know it sounds crazy, but I think most of us 40+ did homework like that?!

Relying on AI to fight AI? The teachers themselves can’t tell the difference? Very soon the tool will start to tell the teachers that genuine human work is produced by chatbots. What about hybrid work: drafted by chatbots, revised by human, can/should the tool tell?

mothafuca 2!

I have tried ChatGPT, any student good enough at asking the right questions should get top marks anyway.

You haven’t convinced me it’s cheating. The use if AI is growing in daily applications and even more so in business. Shouldn’t we be teaching students how to use it?

All homework, except for math, should be abolished.

'New tool that can guess whether students used a calculator on the test'

Can't wait for people to get failing grades based on false positives from this.

Load of crap, eventually some poor innocent is going to get blamed because schools are outdated schmucks relying on black boxes.

😁

How about not having teachers at all. They are unnecessary now..

Bye, bye Touring test. Now we need an AI to tell who is human.

Completely futile. Teachers/school systems have to adapt to the future, not the students.

people have been saying Al is coming to get us for over 200 year's yet here we are still driving our own auto's to the store - Al is just another liberal fantasy

GPTZero may be able to identify whether text was produced by a chatbot - its accuracy is still in question. So it may not be reliable for teachers to determine if students are using AI for their homework. 🤷

The Teachers only have to put assessment rating higher for in class lesson work to assignments and Home Work. In clas 60% other 40%.

It's not like all texts can be resulted from GPT. It depends on the teacher wether she/he makes students write descriptive reasonable/reliable essay or just a simple essay which doesn't need to be corrected

Why are same posts repeatedly posted?

Not for long. If we can make an AI that recognises if something is produced by an AI we can make an AI that creates works that it can’t so recognise.

Students have been copying and hiding the evidence since the beginning of time. This tool though may simply train ChatGPT to be less detectable. AI is not procedural programming, ie if this then do that,

thompkimberly here is an article for you about AI writing

And then, all forget about false positives/negatives... 🫣

Is the seriously advocating teachers to use a closed-source application, written by one person, with no enterprise support, not vetted, and not tested for the ratio of false positives? Imagine all those students having their work falsely flagged as work of an AI?

Who cares? You think this is going away-Nerd are ruining it for everyone

Very similar to my teachers in the 1970s, they just knew who had been copying from someone else's homework.

.

😂😂

Bullshit.

Just use another LLM to paraphrase it and the detector is now useless…

No it can't

They really dont understand 'cheaters'. All they have to do use google to paraphrase it again.

That's so stupid I don't even know where to begin

Trying to stop progress, well good luck with that

What does “high accuracy” mean? 99%? Would that mean 1/100 students will be punished for no reason?

False +ve rate?

This is BS clickbate

The last battling of a collapsing system

everything about wall-e is becoming real

It’s like a classroom drone war.

We have summarized this news so that you can read it quickly. If you are interested in the news, you can read the full text here. Read more:

 /  🏆 541. in TECHNOLOGY

Technology Technology Latest News, Technology Technology Headlines

Similar News:You can also read news stories similar to this one that we have collected from other news sources.

What is ChatGPT and which schools have banned it?OpenAI's latest release of its ongoing artificial intelligence project, ChatGPT, can do some astonishing things and the query-driven system is causing jaws to both drop, and flap, depending on how the output of the the super-smart super-bot is impacting any particular human's day-to-day world.
Source: KSLcom - 🏆 549. / 51 Read more »

The Battle for AI Begins: Vice Defends Political Bias in ChatGPTThe media, which has spent years denying that politically biased censorship took place in Silicon Valley, has now begun denying the political bias of ChatGPT, the most prominent AI chatbot in existence. A war we must win! 😂😂😂 Resistance to sanity.
Source: BreitbartNews - 🏆 610. / 51 Read more »

ChatGPT: What to Know About the New AI TechnologyThe new artificial intelligence-powered chatbot ChatGPT can answer questions, generate essays and even write scientific papers from a short prompt. But its intelligence can only get it so far, often missing the mark on identifying what is true.
Source: wttw - 🏆 520. / 51 Read more »

Google is freaking out about ChatGPTExpect to see a Google search AI chatbot demo this year. Great what now? Ooh this is fun - competition There is a clear difference between searching for Assignment solutions and every other things...
Source: verge - 🏆 94. / 67 Read more »