In the era of artificial intelligence, should assignments and tests at school perhaps be reconsidered?

I agree with the interesting LinkedIn post by Jacopo Perfetti, from Prompt Design, which argues that today there is practically no school assignment or test—essay, summary, analysis, text translation, problem solving, mathematical calculations—that a generative artificial intelligence tool, like ChatGPT, is not able to perform in a few seconds.

And often, even plagiarism detection tools are unable to distinguish between a text written by a student and one generated by artificial intelligence.

But the solution is not to ban the use of AI in schools.

Teachers and educators will have to rethink evaluation criteria: no longer the “final product,” but the process that led to that result:

• encouraging students (from middle school onwards) to use ChatGPT as a starting point, and then analyze, discuss, and argue in class.
• valuing critical thinking, the ability to argue, to ask questions, to delve deeper, and to verify the authority of sources.

In the world that awaits them, what will really matter is not the well-done assignment, but the awareness of the process and the depth of reasoning, competencies that, at least for now, artificial intelligence can only imitate, but not possess.