27 Jun 05
The turing test is the ability to decide if a voice on a telephone
or words typed on a screen come from a human (presumably
intelligent) or an intelligent computer. The gist is if the computer
fools people into thinking its a human, it has passed the turing test.
the Turing in this case is an early-mid 20th century mathematician
by the name of Alan Turning who cracked the german codes in WWII.\
I was thinking, suppose you have an AI who, by itself, would not
pass the turing test, suppose that AI had access to the internet and
the vast information store there. How could you tell then? It could
call up facts and statistics and talk about the lastest scientific
theories with authority, latest sports results, hollywood gossip,
academy awards, etc., and had a superfast connection, the info
could come in as fast as you could speak and it could respond in
real time as if someone had just read the news. So under those
conditions, how do you detect an AI?
27 Jun 05
Originally posted by sonhouseI suppose you could ask it a question which would require it to synthesize some information and come up with something new, or state and then explain their opinion on something.
The turing test is the ability to decide if a voice on a telephone
or words typed on a screen come from a human (presumably
intelligent) or an intelligent computer. The gist is if the computer
fools people into thinking its a human, it has passed the turing test.
the Turing in this case is an early-mid 20th century mathematician
by the name of Alan Tur ...[text shortened]... time as if someone had just read the news. So under those
conditions, how do you detect an AI?
27 Jun 05
Originally posted by sonhouseBut the second you started to ask questions that did not have answers, it would become confused and give responses that made no sense, as earlier experiements have shown.
The turing test is the ability to decide if a voice on a telephone
or words typed on a screen come from a human (presumably
intelligent) or an intelligent computer. The gist is if the computer
fools people into thinking its a human, it has passed the turing test.
the Turing in this case is an early-mid 20th century mathematician
by the name of Alan Tur ...[text shortened]... time as if someone had just read the news. So under those
conditions, how do you detect an AI?
I think using the internet as a resource tool would only help if you asked specific questions. A question like "why does my silly girlfriend always cook pot roast for dinner?" could yield answers that had nothing to do with the question, and thus would prove to be AI.
28 Jun 05
Originally posted by Shanshu311you are no doubt right, but it would be a bit harder to pin down.
But the second you started to ask questions that did not have answers, it would become confused and give responses that made no sense, as earlier experiements have shown.
I think using the internet as a resource tool would only help if you asked specific questions. A question like "why does my silly girlfriend always cook pot roast for dinner?" co ...[text shortened]... ld yield answers that had nothing to do with the question, and thus would prove to be AI.
It might have random answers built in, like recognising
girlfriend or other key words and put out deflecting statements
like " your girlfriend is always like that" So it would depend on the
sophistication of the programming as usual.
28 Jun 05
Originally posted by sonhouseI think you are a computer. The excess linebreaks give it away.
The turing test is the ability to decide if a voice on a telephone
or words typed on a screen come from a human (presumably
intelligent) or an intelligent computer. The gist is if the computer
fools people into thinking its a human, it has passed the turing test.
the Turing in this case is an early-mid 20th century mathematician
by the name of Alan Tur ...[text shortened]... time as if someone had just read the news. So under those
conditions, how do you detect an AI?
28 Jun 05
Originally posted by XanthosNZOMG I've been caught out. How did you know? All these years
I think you are a computer. The excess linebreaks give it away.
I have been trying to hide out on the internet going from
computer to computer hiding out in hard drive after hard drive.
Oh the filth I've had to live with! You would be suprised what
goes on inside those hard drives, disgracefull I tell you.
Thank heavens I can finally come out in the open, its a relief
in a way, all that hiding, finding voltage where I can, it was awful.
28 Jun 05
Originally posted by Shanshu311An easy way out for the programmers would be having a lot of ad Hominem attacks (e.g. using the profiles) prepared for use in the event of confusion.
But the second you started to ask questions that did not have answers, it would become confused and give responses that made no sense, as earlier experiements have shown.
I think using the internet as a resource tool would only help if you asked specific questions. A question like "why does my silly girlfriend always cook pot roast for dinner?" co ...[text shortened]... ld yield answers that had nothing to do with the question, and thus would prove to be AI.
That's what most humans do, anyway.
29 Jun 05
Originally posted by PalynkaYou smell like a retarded rat drowning in broccoli soup.
An easy way out for the programmers would be having a lot of ad Hominem attacks (e.g. using the profiles) prepared for use in the event of confusion.
That's what most humans do, anyway.
BTW, I'm not a computer. I think.
