Do Large Language Models (LLMs) Reason?

One prompt I've found to help "reason" like a human is to tell it to "use common sense" in the prompt

I posted a comment about ChatGPT and it's responses in this thread.

https://news.ycombinator.com/item?id=34870086