TechVIDEO - "I'm 17 for cheating": should ChatGPT be banned in schools?

VIDEO – “I’m 17 for cheating”: should ChatGPT be banned in schools?

This site has the answer to all questions, it can answer in seconds and have a conversation with you for free and in almost any language. The ChatGPT chatbot, launched in November, is a small revolution in the world of new technologies. This tool, created by the American startup OpenAI, is an artificial intelligence, a kind of robot, programmed and trained to communicate in the most human way possible.

Although it arouses the curiosity and admiration of users, ChatGPT worries about national education. Since its inception, it has encouraged students to cheat on exams. A high school student who uses this tool to do his homework for him has agreed to testify anonymously. “The essay question was: “How is the hero responsible for his actions?” I asked him about everything and got 17. I think I’ll still use it for homework, which makes me a little drunk, but it’s true that you don’t have to get used to it.” he explains in the TF1 20H report at the top of this article.

Errors and lack of sources

The Ministry of National Education says it is following this issue very closely, especially since the tool shows its limits as the level of requirements rises. A law professor at the university tested him on very specific questions. And the phrase falls: “He is wrong. I don’t know where he’ll get the answers because that’s still a feature of this software: it doesn’t tell us what its sources are, the hood is closed and we can’t know how to do it.” it works”, worries Bruno Dondero, professor of law at the Sorbonne (Paris 1st).

It must be said that ChatGPT is a conversational robot, “trained” thanks to the phenomenal amount of data gleaned from the Internet, but not “resonating”. Thus, it can sometimes produce a startling mixture of correct answers and factual or logical errors that are more or less difficult to detect. For example, he happens to cite a whale shark (fish) among marine mammals, make mistakes about the size of the countries of Central America, “forget” some historical events, such as the Battle of Amiens in 1870, or invent bibliographic references from scratch.

Answer organized

Faced with the growing use of ChatGPT by middle and high school students or students, some teachers in France have said they are in favor of banning it in schools. A question that also arises in many other countries of the world. In Australia, just weeks after OpenAI provided the tool, eight universities have announced that they are making changes to their exams and believe that students’ use of artificial intelligence is akin to a scam.

Most recently, faced with the viral spread of TikTok videos explaining how to use AI to scam, New York Public Schools restricted access to ChatGPT on their networks and devices. Tool “fails to develop the critical thinking and problem-solving skills that are essential for academic achievement and lifelong success”This was announced by the press secretary of the Department of Education of the American city Jenna Lyle.

However, this opinion is not shared by the entire educational world. Some see ChatGPT as an opportunity to integrate “significant innovation” in programs as well as “calculators or text editors”. Antonio Casili, professor at the Polytechnic Institute of Paris and author Waiting for the robotsbelieves that AI can even help train students’ critical thinking. “For us, this is an opportunity to see how students complete the tasks assigned to them, to make them work on fact checking and to check the correctness of the generated bibliographic references,” it analyzes with AFP.

Others also point to the counter-productive effect of the ban, increasing students’ desire to use the software, as was the case with the advent of Wikipedia or search engines. Finally, the response is organized to better detect texts generated by ChatGPT or any other artificial intelligence. For example, the GPTZero online service is preparing an offer aimed at education professionals, and OpenAI is working on “statistical watermark” set during text generation. Tools that should make writing a task 100% artificial intelligence more difficult.

Editorial TF1info | TF1 Report David de Araujo, Noelle Clerk

Source: TF1


Please enter your comment!
Please enter your name here


Latest article

More article