He can write an essay, answer questions on a paper, or solve a math problem – sometimes with miscalculations. Already used by some students to cheat when rendering homework, the ChatGPT application worries teachers more and more. This tool, which operates using artificial intelligence technology, has only been available since last November. But a French start-up, also specializing in artificial intelligence, has already announced plans to manufacture software for teachers to counter its use by students.
” This software will be ready in the next few months “, assures François-Xavier Hussherr, doctor and associate in economics, one of the creators of “ProfessorBob.ai”, an application developed in 2019 to help students in difficulty to learn better. He has been working on the solutions that artificial intelligence can bring to the field of education for several years. His team has also conducted extensive research to understand how ChatGPT works. The goal? Adapt the algorithms of this future software as well as possible to the problems posed by the American application. ” It will be possible to determine whether an assignment has been completed or not by ChatGPT thanks to algorithms that will analyze the structure of the texts rendered or the vocabulary used in the assignment, for example “, explains the scientist.
A free tool for teachers, with “80 or 90%” reliable results
How will this tool actually work? Teachers will have to go to a website to transcribe the student’s text. After absorbing this data, the machine will be able to give the teacher in a few seconds the probability that an assignment has or has not been written by an artificial intelligence. The results will be reliable, assures François-Xavier Hussherr: “ Even if we can never guarantee a 100% certain result, we will be around 80 or 90% reliable probabilities. This software will be completely free for teachers, promise the designers of “ProfessorBob.ai”.
The innovation may provide a first response to the educational issues related to the development of ChatGPT. Some cases of cheating have recently come to light in several establishments. In December, seven students from the Faculty of Lyon had used this artificial intelligence to write a homework on ” the main features of the medical approach to disability in Europe “. Without ever being identical, the copies nevertheless had a similar construction and arguments. Details that tipped off the proofreader, who then managed to get one of his students to admit to using ChatGPT. All without using artificial intelligence…
We wish to give thanks to the writer of this post for this amazing web content
