Re:Yann LeCun: AI one-percenters seizing power forever is re...
by ryanklee on Thursday November 02, 2023 @11:29AM
(#38114888)
I'm still lost as to how anything you are saying is relevant (and honestly, relevant or not, I'm not following how your sentences hang together).
My claim is that we do have adequate explanations of human behavior and cognition for many purposes. We use them in our everyday life. Without them we couldn't function. We use them in court. Without them we could never ascribe responsibility. And we do all of this without having perfect knowledge of constituent parts. And we find this all acceptable, for the most part, for practical judgement and decision making.
The idea that we must have an exhaustive account of causality in AI whatever to make reasonable assertions about the causality behind is behavior is not true. The reason it is not true is because we do not require this anywhere else in the agentic world.
My claim is that we do have adequate explanations of human behavior and cognition for many purposes. We use them in our everyday life. Without them we couldn't function. We use them in court. Without them we could never ascribe responsibility. And we do all of this without having perfect knowledge of constituent parts. And we find this all acceptable, for the most part, for practical judgement and decision making.
The idea that we must have an exhaustive account of causality in AI whatever to make reasonable assertions about the causality behind is behavior is not true. The reason it is not true is because we do not require this anywhere else in the agentic world.