I was reading stories about a police/stalking case in Las Vegas trying to find the most informative article and came across a very long one that should have been promising but it was so obviously written by AI, and bad AI at that. However, a friend of mine is using AI to do Amazon reviews. lol
I’ll comment that one of my earlier AI-type innovations in assessment was to help prevent universities cheating students, but flagging cases where a tutor’s comments were out of line with their grades. AI is like comedy: it has to punch up, not down, or it is terrible and should be cancelled.
I have a friend that is a professor at ASU and they have been encouraging their students to explore and document their use. However they also teach in a super specific field that AI really can’t complete the assignments.
I can see no risk at all in assigning something which only makes a facsimile of writing without actually making coherent arguments to teach students the basic fundamental skill of writing they need for the rest of their education.
More in class!? When I taught years ago everyone wanted to take the make up test because they could gain access to the questions from friends, frats, etc. I announced that there was no need for an excuse to miss an exam but they would be oral.
Guess what? No more absences.
A lot of schools are taking mixed approaches to AI, where they plan to allow some forms (unclear which, but likely the most profitable ones) and make others punishable, even though AI often creates plagiarism without a source-text so cases are dismissed, and it's a nightmare carousel ride.
We're so dangerously close to any for-profit company/institution being able to run most of the company on AI specifically to avoid having to pay people. When I said this a few years ago people said I was a fucking loon. LOOK WHAT'S HAPPENING, KAREN!
There is already a Seligman-bot trained on Martin Seligman's life's work. And the universities have an archive of materials in Gmail, Google drive, Zoom videos -- suitable for training an LLM. www.politico.com/news/magazin...
and sometimes your university brands its on Generative AI interface instead of partnering with a third party and no one on campus quite knows what to do or how to feel about it www.michigandaily.com/research/umi...
