Summarizing some random text is a quite different task from writing an explainer for a cutting edge AI research paper.

Ah...I don't think this conversation has a future, but I have found that I can use an LLM to give a pretty damn good summary of some fairly verbose and well-organized "random text."