But, this doesn't work when our students lack the confidence to question the AIs (and who can blame them given the amount of marketting the tech companies are doing)

The AIs can give good advice on how they should be used:

While this isn't a huge surprise, what I find despressing is how confidently wrong the AIs still are. When asked, they even admit that they're wrong, and that they are likely damaging educational outcomes.

Interesting to note that Claude (Sonnet 4.5) gets the Clarke & Hunt (2016, Psych Science, journals.sagepub.com/doi/abs/10.1...) Focus-Divide Dilema wrong.

I'm starting to wonder why the UK's HE sector thinks the government should bail the sector out? Seems like a waste of effort given the 0% attendance.

It's a shame the students don't want to turn up as I enjoy teaching.

"For all experiments, we have reported all measures, conditions, data exclusions, and how we determined our sample sizes. Data and materials are available on the Open Science Framework osf.io/xxxxx).%22

...rolls eyes...