[ExI] Trilemma of Consciousness

[Stuart LaForge]

John Clark wrote:
​
>1) Strictly speaking the Turing​ Test is a test for intelligence not
>consciousness, although until AI started making big gains nearly everybody
>made the unspoken assumption that the two were linked, but now many want to
>believe that carbon atoms are conscious but silicon atoms are not.

Yeah, your definitely right about this. From the context of biology, it
started back in the Darwin days. As people realized that animals could be
smart too, they started raising the bar from intelligence to consciousness
without really defining either,

>2)  The Turing Test has nothing to do with Turing Machines, the test is
>agnostic as to how the subject manages to produce the observed behavior,
>it's irrelevant.

Yes. That is why I call it the generalized Turing test (GTT) to
distinguish it from the historical Turing test which started out as a
party game for participants to guess the sex of opposing players through
typewritten notes to one another. I could just as easily called it the
generalized Voight-Kampff test but as much I am a fan of Phillip K Dick,
that bloke Turing did save us all from Nazism.

>3) The Turing Test is not a mathematical theorem that can be proved or
>disproved, it's a rule of thumb, it's just an acknowledgment that
behavior >is the only way to judge the intelligence of others.​

Yes. I have done neither. I have instead framed it as a decision problem
and bounded its results. Behavior is indeed all that matters, but it
doesn't hurt to thank the vacuum cleaner too. After all, purpose is a
semantic property also.

​>A​nd we can make mistakes and judge a stupid being as smart as in the
>character of Chauncey Gardner in the movie "Being There"; another example
>would be the millions of voters who judged Donald Trump to be smart back
in
>November. And of course it would be easy for a smart being to pretend to be
>stupid and fool us. The test isn't perfect but it's all we have.

And all we'll need . . . I hope. :-)

Stuart LaForge