Of course, we're not there yet, but if our students manage to use the tool correctly, even if it's just to clear some of the groundwork, that will already be very good. We can envisage various educational scenarios: they could use it for course summaries, mind mapping, redefining certain concepts, etc. However, I remain skeptical and cautious because we are still in the early stages. I submitted three physics problems to ChatGPT, and it got two of them wrong... However, the goal is not to "trick" ChatGPT: that would be to treat it as a rival and forget that it is only a machine that is destined to improve. "If I use a hammer, I don't compete with the hammer..."
Prof. Michaël Lobet, Faculty of Science, Department of Physics, NISM Institute