I'm not uncritically supportive of the suggestions offered here, but Philippa Hardman has summarized recent work on cognitive offloading in AI use and comes up with recommendations much more nuanced than a simplistic 'ban AI' response. If I had to summarize, I'd say that the advice amounts to saying: let AI do the stuff that the learner needs (explaining new concepts, performing rote work), use it as a partner where necessary to free capacity, but critically, ensure "the freed capacity is reinvested with intention." So there's no less effort on the part of the learner, but rather, the effort is directed toward what they really need to learn. Finally, keep authority with the learner. These are the findings; the six 'principles' that follow are to my mind much weaker, expressed as generalizations without the specificity found in the earlier sections.
Today: Total: [Direct link] [Share]
