PDA

View Full Version : Artificial Intelligence



Matthew
2003-Dec-06, 05:03 AM
When will we humans create artificial intelligence? Check out current AI technology. (http://www.a-i.com) On www.a-i.com (http://www.a-i.com/) you can talk to the Alan AI Bot, or to one of his many little brothers; the many personalities of HAL.

Often when you talk to the Alan you can't tell that it isn't human. Its truly wonderous, and it makes you wonder at what is to come with more advanced software and more powerful hardware.

Haglund
2003-Dec-06, 07:20 PM
I agree that Alan was one of the most wellmade bots I've seen so far. I've chatted with a few :) (even tried to make a few myself, but never got very far)
Depending on how we define intelligence, one might say we already have AI, although limited. What is intelligence? Must an intelligent entity be aware, or selflearning? Or is idle information processing enough to qualify for being intelligent?

DippyHippy
2003-Dec-06, 11:33 PM
I tried Alan but found the conversation to be a bit boring LOL Actually, I thought he was quite self absorbed... he only wanted to talk about himself and how he and HAL works LOL

I couldn't get HAL to work either... got a list of personalities up, but when you clicked on them all they did was give a description of how they worked.

Truthfully, I was impressed - I couldn't write anything like this - but I'll be really impressed when they create something and you can't tell it's a program. Even if you tried to claim that Alan was a real person, I wouldn't be convinced but I don't think we're far off from that level of reality.

Actually, I'll be extremely impressed when they invent Teddy out of AI :lol: I remember talking about the film with a few people and we all agreed we didn't really care what happened to David and his blue fairy quest, we just wanted to make sure Teddy got it through it all okay LOL

Matthew
2003-Dec-07, 01:30 AM
If you want to talk to a HAL personality just say "I'd like to talk to (the HAL personality you wish to talk to)."


Depending on how we define intelligence, one might say we already have AI, although limited. What is intelligence? Must an intelligent entity be aware, or selflearning? Or is idle information processing enough to qualify for being intelligent?

AI should be able to learn, it should also be aware that it exists. But if it became aware that it existed couldn't you face something that wanted to control. You could get something like what happened in Terminator 3, the AI controlling the whole internet. Which would be pretty scary.

Haglund
2003-Dec-07, 10:37 PM
Originally posted by matthew@Dec 7 2003, 01:30 AM

Depending on how we define intelligence, one might say we already have AI, although limited. What is intelligence? Must an intelligent entity be aware, or selflearning? Or is idle information processing enough to qualify for being intelligent?

AI should be able to learn, it should also be aware that it exists. But if it became aware that it existed couldn't you face something that wanted to control. You could get something like what happened in Terminator 3, the AI controlling the whole internet. Which would be pretty scary.
Selflearning should be possible, the more difficult thing is selfawareness. When is the AI aware of itself and/or its surroundings? And, how could it become aware? What IS awareness? Is the flesheating plant aware of the insect that it capture and digest, or is the ant aware, and how much, does it know it's building a colony, for example? One interesting approach to AI is to integrate it into a humanoid robot, because then it can get to interact with the environment. I mean, Alan doesn't know what the conversations are all about, really (but if it perfectly simulate human behaviour, does it really matter? The black box thing and all (that you know what a certain function of a program does, but not necessarily how.)). Anyway, humanoid (or, for that matter, other types of) robots is a great idea. The robot can see, hear, touch and learn more. One interesting example is the COG robot over at MIT (http://www.ai.mit.edu/projects/humanoid-robotics-group/cog/cog.html). Will such robots achieve a level of awareness that we have, and if so, what will happen? But, then again, with the right software maybe a PC could be "aware", where it can interact with the surrounding through the keyboard and monitor... perhaps. Just a few thoughts... it's too late and it might show in my post, hehe.

DippyHippy
2003-Dec-08, 12:50 AM
AHhhhhhhhhhhhhhhhhhh... I didn't realise that... I just typed "I want to talk to HAL" and then it was giving me the list of personalities....

...duh... no intelligence here then - artificial or otherwise :D

Chook
2003-Dec-08, 07:56 PM
Wasn't it the Englishman ,TURING, who did a lot of thinking in this field? I think he defined AI and developed a "Turing Machine" - something like the HAL environment you are talking about, as I remember.
(The poor begger later blew his brains out when his homosexual "preferences" were exposed and he fell from grace.)

eggplant
2003-Dec-09, 04:32 AM
Do any of these bots get jokes? As a victim of a couple (in odd chat rooms) I gotta wonder...

Haglund
2003-Dec-09, 12:19 PM
Originally posted by Chook+Dec 8 2003, 07:56 PM--></div><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td>QUOTE (Chook &#064; Dec 8 2003, 07:56 PM)</td></tr><tr><td id='QUOTE'>Wasn&#39;t it the Englishman ,TURING, who did a lot of thinking in this field? I think he defined AI and developed a "Turing Machine" - something like the HAL environment you are talking about, as I remember.
(The poor begger later blew his brains out when his homosexual "preferences" were exposed and he fell from grace.)[/b]

Yes he was one of the big ones in the field of computer science. He was interested in AI and wrote about the Turing test. Also participated in the attempts at cracking the german Enigma encryption.

Here is the paper Computer Machinery and Intelligence (http://www.loebner.net/Prizef/TuringArticle.html) where the Turing test is described.

<!--QuoteBegin-eggplant@Dec 9 2003, 04:32 AM
Do any of these bots get jokes? As a victim of a couple (in odd chat rooms) I gotta wonder...[/quote]
Didn&#39;t try it out, but I doubt it. Some of them tell jokes, though.

Chook
2003-Dec-10, 08:43 AM
Thanks a lot PARKER - I had a look at "Computer Machinery and Intelligence" and found it fascinating, but very dated, stuff (horse and cart days).

I guess we&#39;ll never get human emotions programmed unless we say something like - "This situation would normally provoke anger in a human - which would result in THIS reaction ..."

I remember early UN computer-translation days - specifically RUSSIAN-ENGLISH/ENGLISH-RUSSIAN. The problem was, for example, how would you translate "spirit" within a sentence? It is situation-dependant i.e "The SPIRIT of the meeting was cordial." verses "Because of the ambiant temperature some guests prefered SPIRIT(S)" - so a preliminary analysis of each sentence had to be made before proceeding further.

Yes - there have been computer-generated jokes based on the science of the analysis of a joke. Actually I have read that Bob Hope used these services.