Yesterday I prompted three different chatbots (ChatGPT, Bard, Poe) with a prompt to generate lyrics on a specific topic (VC investment in Europe) in the style of the Gaston song from Beauty and the Beast." They were all pretty terrible. Then I installed the Bing app that comes with GPT-4 and it did a fantastic job, so there's that. I don't pay for OpenAI's GPT-4, so can't compare that one.
I recently asked Claude+ a social science question (nothing controversial) and it answered much better than ChatGPT. Chatgpt didn't want to cite studies and seemed uncomfortable even being asked about the topic and deflected while Claude seemed to genuinely try to answer the question.
