There is an urgent need to move from an instrumental use of AI to new didactics that problematize the context, accompanied by a focus on building learning communities.

It was in November 2022 that the new version of OpenAI’s GPT-3.5 language model timidly appeared on the scene. Soon many analysts began to see the potential of this tool and the enormous changes that the new version of the well-known artificial intelligence showed. However, the bombshell exploded when the same company launched version 4.0 on 14 March 2023, triggering a race that has yet to be halted by the major players in the field.

By Alexandre Gattreaux<

A year has passed since that milestone, and the voices of those who predicted the end of thousands of jobs and, in the world of education, the early end of English and history teaching, among others, are still echoing. How can we forget those university institutions in the US and Australia that have blocked these tools on their campuses or reverted to paper-and-pencil assessments?

What is certain is that today we have a different panorama from the apocalypse predicted by supposed experts. Schools continue to function in the same way, and the truth is that students and teachers are now incorporating this tool into their activities.

A survey by educational technology company RM Technology found that two-thirds of teachers believe they regularly receive work written by AI, while 49% of students surveyed believe that not using AI would be detrimental to their learning. These figures are in addition to those from the World Economic Forum, which shows that the percentage of tasks performed by automation is expected to rise from 34% in 2023 to 43% in 2027.

The question, then, is how teachers should use this new technology. Experience to date shows that it is being used slowly and mostly to speed up administrative tasks or to repeat the same pedagogical actions that have little impact on learning.

There is therefore an urgent need to move from an instrumental use of AI to one that allows for new didactics that problematize the context, accompanied by a focus on building learning communities.

It is in this narrower and more focused working space that AI cannot operate, even more so if the work is based on pedagogical engineering that emphasizes problematization rather than replication of a standardized curriculum.

In this way, it is understood that the act of teaching is far from being replaced by a robotic process, as AI has so far demonstrated a serious lack of deep contextual understanding and human empathy. Moreover, pedagogical intuition, a key feature of human teaching, may be difficult to replicate with AI.

This technology also shows a significant lack of ability to promote social and collaborative learning, which is crucial to the educational process, and finally, an inability to stimulate the development of critical thinking.

We are therefore called upon to review our pedagogical practices to face these new challenges, to make a turn in our didactic work, and to integrate the use of AI in a mediated way to promote meaningful and situated learning based on reading the context, using active methodologies that emphasize the construction of a learning community.

Alexandre Gattreaux, UMCE academic expert in innovation, didactics, and the use of AI in education.