pd (02/05/83)
I will take the risk of repeating what I had argued. I think I might have made some observations that are not strictly relevant to the main thrust of my arguement. The crucial point is not that "One can not prove or disprove that a machine has conciousness". The crux of the arguement is that conciousness is a non-linguistic, non mathematical, non dualistic experience. Such being the case, it is not possible to describe it any way, and hence it is impossible to build a machine that has conciousness. The main stumbling block that I am identifying is not one of validation, ie., ensuring that claims of machine conciousness are correct, but rather one of modus operandi... I can't see how some can go about describing conciousness, modeling it, developing a mathematical notation, a physical design.. etc: steps that are normally involved in building any kind of machine. By the way, the only time my friends are not concious is when they are drunk. Prem
rostain (02/07/83)
You seem to be making a basic mistake, in your defense of yourself. Just because one can't prove that consciousness exists, doesn't mean that it doesn't exist, or is less real. I agree that we can't build a machine that we know has consciousness, but maybe we can build one that we believe has consciousness? I think so, as long as we see it's behavior and actions in terms of human behavior and human actions (Turing Test?!). Since consciousness can't be proven or disproven, the distinction between believing something is conscious and knowing it is becomes unimportant. Thanks, Alain Rostain decvax!yale-comix!rostain