In a letter to teachers, Sergei Guriev, director of education and research at the Institute for Political Studies (IEP), announced that his students have been banned from using ChatGPT, the conversational AI developed by OpenAI.
Teachers are concerned that ChatGPT is being used to cheat
Unveiled in December, ChatGPT is based on the GPT-3.5 language model. This artificial intelligence is capable of maintaining a conversation like a human being, as well as recognizing mistakes. It is shaking up the tech sector as its performance is so impressive that OpenAI has seen its valuation evolve exponentially since the launch of the chatbot.
“ This tool, which uses artificial intelligence (AI), strongly interrogates education and research actors from all over the world on the topic of fraud in general and plagiarism in particular. Some states have already banned its use in their schools and universities. writes Sergei Guriev, referring here to New York schools that have banned its use.
The IEP, commonly known as Sciences Po, has decided to go the same route: ” The use of ChatGPT at Sciences Po, or any other tool using artificial intelligence, is, with the exception of pedagogical use supervised by a teacher, strictly prohibited for the moment during the production of written or oral works by the students under penalty of sanctions that can go as far as exclusion from the institution or even from higher education » continues the director.
A revolutionary model, but one that worries
It also states that ChatGPT issues “ it will inevitably and rapidly change teaching practices and teaching evaluations », and that a note on the subject will soon be distributed to his teachers.
For its part, OpenAI is already working on the GPT-4 model, which should be even more powerful than ChatGPT. The company would also like to charge for its conversational AI. Microsoft, which has already invested in the company, announced additional multibillion-dollar funding in OpenAI, as well as integrating the tool with its suite of Azure OpenAI services.
Language processing models are increasingly sophisticated and are revolutionizing many fields. However, they may be plagued by racist or sexist biases or spread false information. For this, experts are calling for the development of powerful tools capable of supervising the exploitation of systems such as ChatGPT.
#Sciences #prohibits #ChatGPT #students
0 Comments