There's no doubt that generally LLMs are better. In addition SO had its issues. That being said I can't help but worry about losing humans asking questions and

There's no doubt that generally LLMs are better. In addition SO had its issues. That being said I can't help but worry about losing humans asking questions and humans answering questions. The sentimentality aside, if humans aren't posing questions and if humans aren't recommending answers, what are the models going to use?