An amusing reflection into a modern day social hierarchy according to the public online sphere. Also fun to imagine that most of these language models are snapshots of consumable and parseable discourse and so their bias will be a time capsule of the overton window from a certain time period.
That's the perspective I take as well: LLM's are giving humanity (imperfect) insight into it's collective consciousness (how they describe their experience and perception of reality (aka: "reality", "the reality", etc) in text), almost like holding up a giant mirror to ourselves.
I'm thinking before too long this idea is going to leak out into the collective consciousness, and that may make things in this already utterly bizarro world even more interesting than they already are. It's gonna be a wild decade!
Exactly. What's interesting as well is how much this initial point in time will perpetuate forward as future language models are built on the output of the current ones, dampening any potential shifting in the overton window. They may try to avoid ingesting their output (ChatGPT says it can detect use of itself), but as more independent models get developed, the detection will never be perfect and in fact may be impossible at some point.
