I read this post: https://technicalwriting.dev/data/embeddings.html and its discussion on Hacker News https://news.ycombinator.com/item?id=42013762 and think it seems really interesting. I have never heard of so-called Embeddings before but could instantly feel the itch to start analyzing and auto categorizing text in bulk. So much so that I didn't want to forget the technology, so that's why I'm writing it here. There are probably hundreds of applications for this, but one that instantly came to mind is doing some analysis on news articles to identify visually what news outlets write about. Could embeddings be used to determine political angles for example?