By Fabienne Serina-Karsky, Catholic Institute of Paris (ICP) and Gabriel Maes, Catholic Institute of Paris (ICP)
The introduction of a new technology usually raises very clear-cut reactions, between enthusiastic welcomes and stubborn reticence. Artificial Intelligence (AI) is no exception to the rule and raises dilemmas. However, rather than asking whether it should be welcomed or banned from the educational system, should we not first of all start from the postulate ofeducability We must ask ourselves how to accompany each learner towards the learning that is necessary to find his or her place in tomorrow's world.
So let's ask the question again, "how can we integrate AI into our school curricula to better support each learner toward their own unique excellence?" Given the constant evolution of technology, it is essential to weigh the concerns raised by the use of AI, but also to see how it could promote access to quality education by reinterpreting pedagogical practices and the posture of the teacher, thus becoming part of a revolution in teaching methods more than a century old.
The emergence of artificial intelligence (AI) in the education system is a double-edged sword. On the one hand, AI can be used as an effective tool to help students by providing personalized learning materials and instant feedback on their work. On the other hand, if AI is used without care, it can have detrimental effects on their performance and preparation for the job market.
The recent example of ChatGPT shows us that some institutions have banned it for fear that it will facilitate cheating and lower academic standards, while others have welcomed it into their classrooms because they feel it is impossible to fight these technologies - and fight a losing battle. Both positions are defensible. Institutions that have blocked access to these technologies are well aware that if a learner's work is relegated to 100 % to the technology, there is little or no learning, especially at a key time in adolescent development.
However, this prohibition proved to be relatively ineffective, as students easily bypassed it by using the 4G connections of their cell phones, thus giving reason to the supporters of their integration. How, then, can we best support learners, but also teachers, in the use of AI?
The European Commission has been thinking about the integration of digital in education for several years, from elementary to higher education. To ensure that teachers and students benefit from the potential of AI for learning, it published guidelines for the ethical use of AI and data in education and, on September 30, 2020, approved the digital education action plan 2021-2027. These guidelines are intended to provide support at all levels - from instruction to the administrative tasks associated with it - so that everyone can have an optimal learning experience.
In North America, the National Federation of Teachers of Quebec gives us an example of how to deal with the issue.
In France, at the end of January 2022, the Minister of National Education Pap Ndiaye presented a digital strategy for education 2023-2027, the aim of which is to strengthen the digital skills of students and accelerate the use of digital tools for student success. The different axes and measures are presented in a 41-page report that develops in particular the points of a "reasoned, sustainable and inclusive" digital offer for the benefit of an educational community and to "enable students to become enlightened citizens in the digital age".
The reflection undertaken goes far beyond the framework of the school or university and requires the mobilization of all educators, parents, teachers, to accompany the new generations in the use of these technologies that are disrupting the educational uses especially in the context of the school and university.
The ChatGPT turning point
ChatGPT technology, which has been making headlines since its launch on November 30, 2022, is now common knowledge. Many middle school and high school students have tried to delegate their homework to AI. And many teachers have reported their helplessness, as it is so difficult to detect whether an essay is written by a student or by the AI, especially since it is possible to ask ChatGPT to adapt the essay according to the status of the student, whether it is a middle schooler or a student preparing a thesis for example.
OpenAI, the company that designed ChatGPT, and others, promised to produce scripts that would differentiate between human and AI writing. Today, it works quite well on English texts. However, on the ten or so tests we have done on French texts with plagiarism detection software, the results show that, on texts written by an AI, in 60 % of the cases, the software does detect the AI and therefore in 40 % of the cases, it thinks it is a human being.
On the other hand, we also found that for AI-generated texts, it was enough to replace two or three words in each sentence for the testing software to think that the writing was human. The only alternatives that work so far and that teachers are forced to use are tabletop written assessments without Internet access and oral assessments. However, not all universities have the technical and/or logistical means to organize all face-to-face exams. And the evolution of AIs is exponential.
These transformations question the evaluation of skills and could lead to the credibility of diplomas being undermined. It is conceivable that recruiters will no longer be satisfied with academic credentials and will add diagnostic tests to verify the skills claimed by the candidate's CV. This would encourage students to focus on skill acquisition and stop focusing on grades. Would the development of AIs encourage us to think differently about school?
A Copernican revolution
The Swiss psychologist Édouard Claparède, from the beginning of the XXe century, speaks of initiating a Copernican revolution to recognize the child's capacity to be an actor in his or her own education. The educator would then no longer be a "teacher" but an "informant", to use the words of Roger Cousinet, a French inspector who participated, in particular, with the famous Maria montessoriIn 1921, he became a member of the international New Education movement, which was federated with the aim of transforming education.
Pedagogical innovation is then invited to the school through different tools and methods no longer based on a masterly and identical teaching for all, but on a learning based on the specific abilities of each student. This is the inclusive school before its time. By starting from what makes sense for the child or young person, the teacher brings him the elements he needs to build his project, and thinks of evaluation differently.
However, in order to face the risks of new technologies and the avalanche of information now available to all, rethinking the role of the teacher seems to be one of the key factors. Added to this is the challenge for tomorrow's school to integrate the new knowledge needed for the education of the futureThe philosopher Edgar Morin identifies knowledge, uncertainty and error as key elements.
Faced with the mass of freely available knowledge and with AI that now allows it to be used more or less wisely, education in the search for information and its proper use is an opportunity to make learners aware of the meaning that everyone can derive from learning. In conclusion, it seems urgent to develop critical thinking and to ask ourselves how the school can seize these new challenges to continue this Copernican revolution by relying on the new tools that are sure to appear.
By Fabienne Serina-KarskyHDR lecturer in Education Sciences, Catholic Institute of Paris (ICP) and Gabriel Maesa pedagogical engineer and trainer, Catholic Institute of Paris (ICP)
This article is republished from The Conversation under a Creative Commons license. Read theoriginal article.