From the Digital Education newsletter.
Leaving aside issues of copyright and related matters, two things strike me about this question. Firstly, if students can cheat by using AI to complete their assignments. does that not imply that the assignments were not fit for purpose in the first place (assuming the purpose was to assess students’ knowledge and understanding of the particular topic?) The challenge, surely, is to devise assignments that require more than answers based on probability.
For example, one thing that Al is pretty good at, notwithstanding its hallucinations, is searching the web and even documents on the web in some circumstances. So getting students to do the research in order to do something with the results, rather than getting AI to do the grunt work, seems to me like some teachers I’ve observed getting their students to enter data into a spreadsheet in order to carry out calculations and predictions. What tends to happen is that by the end of the lesson some students have not even finished entering the data. It’s a complete waste of everybody’s time. Much better to furnish students with a populated spreadsheet so that they can focus on what sort of formulae to insert in order to complete the task.
By analogy, I think getting AI to do the heavy lifting in terms of research makes sense. Yes, they will have to spend a bit a time evaluating AI’s results, but in my opinion it would still be a better use of time.
Or perhaps, as a compromise, they could carry out the research in the old way (!), or write down what they think might be the points that come up, and then use AI as a check. That’s what I did with my piece on objections to PISA (see below). I made a note of what was wrong with it in my opinion, and then asked At what the objections to it were. (I’m relieved to report that it came up with the same list!)
The other thing that strikes me about the question of whether it matters if students use AI is that it is similar to a question I liked to put to ed tech coordinators when Microsoft introduced wizards, My question was: if a student uses a wizard to, say, produce a leaflet, rather than create it themselves, does that suggest that they have a high level a digital literacy or a low level? It always led to a lively debate, but my own view was that getting a wizard to get something in place that you could then amend and adjust was a much better use of time than painstakingly creating it from scratch.