Any discussion of LLM‘s will eventually include some reference to training. Positive aspects as well as the negatives always seem to point back to how the LLM was trained. I’ve been pondering how that is different from how humans are “trained.” Eventually, I suppose a human child will take over its own learning experience. But isn’t that likely where we’re headed with LLM‘s?

I put that question to ChatGPT and, as so often happens, a lengthy discussion ensued, touching on what it means for an LLM to “trust” and should human trust be considered superior, which led to the concepts of self and free will.