Lindie Koorts wrote on Facebook:
Just completed my first round of marking undergraduate essays in the age of ChatGPT et al.
1) If an essay is good, one distrusts it straight away. I did not award a single distinction.
2) The really good essays do not address the essay topic directly, and instead go in for generalised descriptions. Dead giveaway.
3) Many students use free AI-software to generate synonymns in order to throw off plagiarism- or AI-detection software, which results in a garbled, incomprehensible mess.
4) I ended up passing essays that I would have failed in the past, simply because I could tell that the students had actually attempted to do the work themself.
5) It is a strange world we live in when you place more trust in mediocrity than quality, and where you reward accordingly.