[jimmyjazz14]

I would like to see the concept of what an LLM is move away from its awkward chatbot phase and more into an era of utilitarian functionality (which is where they really shine anyway). The problem I see it is that LLMs got anthropomorphized early on (which was probably inevitable) so people actually believed the AI was thinking about their problem and considering it when it really wasn't, if we just thought of them as really good auto-complete engines, or better search engines, it would matter less what the LLMs sentiment was towards the users (as it probably shouldn't have any).