I wish I shared OOP's optimism.
None of the people I talk to offline can shut up about how LLMs are actually fine if they're used as a tool by experts in the field, and how none of the ethical problems people point to are inherent to the technology as a whole and therefore people who criticize LLMs because of those problems are overzealous idiots. Some even say that none of the ethical problems are that bad when presented in isolation from each other (or worse, that they aren't against the law and are therefore okay), and they definitely don't apply if you use a local LLM (which often isn't really true, plus it's a 50/50 chance that the person saying it uses Claude). Most on the Fediverse are united in their hatred of AI, but lately I've seen several people on here (both on Lemmy and on Mastodon) start shilling for it too.