All the hand wringing about openAI cheating suggests a question: why so much mistrust?
My guess would be that the persona of the openAI team on platforms like Twitter is very cliquey. This, I think, naturally leads to mistrust. A clique feels more likely to cheat than some other sort of group.
I wrote about this last year. The levels of trust people have in companies working in AI is notably low: https://simonwillison.net/2023/Dec/14/ai-trust-crisis/
My take on this is that people tend to be afraid of what they can't understand or explain. To do away with that feeling, they just say 'it can't reason'. While nobody on earth can put a finger on what reasoning is, other than that it is a human trait.