[originalcopy]

I'd like to see some examples of when it struggles to do summaries. There were no real examples in the text, besides one hypothetical which ChatGPT made up.

I think LLMs do great summaries. I am not able to come up with anything where I could criticize it and say "any human would come up with a better summary". Are my tasks not "truly novel"? Well, then I am not able, as a human, to come up with anything novel either.