350M TOKENS DON'T LIE: LOVE AND HATE IN HACKER NEWS

Summary from the TLDR Web Dev newsletter

This post analyzes Hacker News posts to understand community sentiment and trends. Using a large language model, its author analyzed over 100,000 posts and millions of comments to identify popular topics, like programming and computer science, and topics that evoke negative reactions, such as FTX and police misconduct. Certain topics, like GNOME and Google, can generate both positive and negative reactions because they are divisive.