2226
Living Room / Re: The Story of Merlin: The 1978 Electronic Game
« Last post by 40hz on March 06, 2014, 12:13 PM »^I don't think I'm ... giving undue emphasis to what passes for human intelligence....-40hz (March 06, 2014, 05:42 AM)
@Tao = You completely lost me. ... "No True Scotsman??? Love it! But I have absolutely no clue as to what that is alluding to!![]()
)
-TaoPhoenix (March 06, 2014, 08:59 AM)
Put another way:
As I understand it, the Turing Test was supposed to be about a person talking through one terminal and a computer program talking via another one, and the human operator is supposed to try to figure out which is which.
But I think there are some assumptions going on about the level of intelligence of the participants. So if the human on the other side of the terminal is less coherent than a chatterbot and can't type either, and the test taker is also feebleminded, and if the chatterbot is tuned well, it very well could win!
So the "No True Scotsman" part kicks in if we start trying to say things like "oh, well, that's not a true test...".
Ah...ok. I got confused because I was talking about Turing's Halting Problem, not the so-called 'Turing Test.'
FWIW, the Turing Test was intended to be wholly subjective - which makes the validity of the result entirely dependent on the person evaluating the responses received. Hardly "scientific" in the modern sense. But as to whether or not the Turing Test is a "true test," I have no idea since I'm not 100% sure what I think 'true' is in this domain - or how it could be established beyond all doubt. (Thank you Kurt Gödel!) So I tend not to want to use the Turing Test as an bullet point in discussions of human/machine intelligence (either pro or con) because it's such a hokey test anyway.
Sorry for the misunderstanding - and thank you for that "No true Scotsman" phrase. I love it!


Recent Posts
)

Modula!
And Oberon! 

