Does Artificial Intelligence make skills worse?
Need to continue to keep the focus on a conscious, careful and transparent use of technology
2' min read
2' min read
From the outset, the spread of generative artificial intelligence tools has raised concerns and doubts about how the ability to instantly create sensible and relevant texts could undermine certain pillars of the educational system.
Data from the study Generative AI Can Harm Learning, which appeared in 'The Wharton School Research Paper', seem to prove that these concerns are well-founded: the study was conducted in a school setting and shows how access to Gpt-4 (one of the most popular AIs) improves student performance, but reduces autonomy when these same tools are missing.
The research, conducted on high school students, showed that the use of Gpt-4 led to better than average results in mathematics tests. However, when access to Ai was withdrawn, students who had used it performed worse than those who had never used it. The only exception was found with the use of Gpt Tutor, a version of Ai equipped with mechanisms specifically created to support learning.
Generative Artificial Intelligence, in short, makes it possible to improve human productivity, but raises broad questions about the consequences it may have in relation to learning. The risk, conclude the authors of the paper, is that Ai becomes a sort of 'crutch', hindering the development of the skills needed to tackle complex problems without external help. The risk is definitely higher when we are talking about school age, i.e. a crucial phase of personal development.
What do these data tell us about the future of education?
