turing

[iain]

we are currently looking at Turing in CSC lectures, has anyone taking a turing test?

[Fool]

What is it?

[DrakkenKorin]

Only Turing I've ever heard of was Alan M. Turing, the one who helped decipher the German Enigma codes.

That the one?

[EDIT]
Found some info and a link:
http://www.dictionary.com/cgi-bin/di...=Turing%20test
[/EDIT]

[almost_here]

The Turing Test is not something people take

Turing believed that the only way to prove an artificial 'thing' had intelligence - or AI - was if it could convince a human being it was real in that sense. It's highly disputable, but he said a computer would only be intelligent when it's responses to questioning indicated to a human being that it was so.

Ah.

[mithrandir]

That sounds facinating. I'll be doing Decision Support Systems next year so who know's? I may actually read more about Turing.

[adrianxw]

Interesting gut. A lot of AI theory came out of his writings. Frankly, I do not believe the Turing test is a true test of intelligence, as I have said before.

I gave the example last time of the dolphin. Nobody questions the intelligence of a dolphin, but a dolphin could not pass the Turing test. Is it intelligent enough, don't know, we are not intelligent enough to explain to it what it has to do!

[iain]

>>its not a test you can take
that line was a sarcastic question from the Hacker Test

But i was just wondering if anyone had any views on it, and does anyone belive that we will acheive totally artificially Intelligent*
machines.

*for the sake of argument - i am assuming an AI machine to be one capable of action in response to processing of the surroundings and facts available and a system that is capale of learning. However if anyone would like to challenge this defintion please do

[adrianxw]

Hmmmph! I wish you'd been in the other recent AI thread. There, everyone wanted AI to equal human! I'd have had at least you on my side! I think your definition of AI is much closer to mine than their's, (see the dolphin example I gave).

[novacain]

>>i am assuming an AI machine to be one capable of action in response to processing of the surroundings and facts available

The system I am working on / almost finished sits idle for 99% of the day. When a 'event' happens by it wakes up and photographs it. When the 'event' is over it shuts down and processes the images.

>>and a system that is capale of learning

It then collates the data and validates further based on it's database. Allowing it to suggest errors, replacements and trend wear of components. Finally addding to the database for the next event.

I don't think it is anywhere near an AI, but it is comming close to your interpretation.

Yes, adrianxw, the Turing is not an accurate test in my opinion. We should not be trying to re-build humans (flawed concept the human). But it does test abstract thought which is its point.

Got a better test?

My understanding is the best entry in the last tests would sing some nursry rhyme if it could not understand the question / conversation.

[adrianxw]

>>> Got a better test?

I don't think the problem lies with the test per se, I believe the problem lies with the definition of AI. If you consider iain's intelligence points, that much better applies to a dolphin then a turing test candidate system!

What is intelligence?

[VirtualAce]

I've seen way to many Terminator type movies. Computers are getting larger (memory wise) and faster but I hope they never have sentient AI. The world is bad enough with human AI running it. What would a computer do? So when it comes to this topic - I like stupid computers. Plus, I can always turn the thing off if it goes wacky.