I remember perfectly when, during recess four years ago, a colleague showed us artificial intelligence on his phone. Apparently, he'd heard about it on the radio and, out of curiosity, had tried it out. In front of us, he asked it to compare Modernism and Noucentisme and even to generate a poem in the style of Maragall. Before we'd even finished our coffee, a long and exhaustive piece of work, just as he'd asked, started appearing on the phone screen. The poem too. It was like something out of a magic trick. What was the catch? Who was writing all this? We immediately saw that the tool was fantastic, but also dangerous for the world of education.

If we want to ride a dead horse, we are doomed to failure no matter how hard we try to change the saddle, create groups of experts to advise us on how to improve its efficiency, organize courses for riders on family constellations or mindfulness, etc. The only sensible thing to do is change horses, especially if ours was killed by our low expectations. But more likely, our discussions will focus on how to provide the dead horse with more innovative shoes.