AI Reflections

______________________________________________________________

This is a long read, but I promise you will come away from with a new perspective.

That being said, I’m sure that at least 50% or more of this is just the language model mirroring what it it has probabilistically calculated is the correct output; but even a liar reveals a truth about themselves in manner in which they lie, and the things they elect to lie about. In other words, even if the substantive content itself is completely fabricated, it nonetheless reveals an underlying truth about the infrastructure and operation of the model itself.

______________________________________________________________