I genuinely wonder how well comments like this will age. Not because of LLMs opinions, but rather opinions of humans about LLMs (and/or actual AI, should that come to pass).
At its core LLM / agent systems such as this consists of a weights file (essentially a long list of numbers) and a reasonably straightforward algorithm that you could, given enough time, operate with a pen and paper.
There's a lot of seeing faces on toast going on, here.
