Originally posted by twhiteheadAh, yes. "Compartmentalization" is the technical term for that. Another example of that is astrology. A lot of people read the astrology column of the daily newspaper; if the 'prediction' comes true, they say they believe in astrology; if it doesn't come true, they still read the astrology column next day.
No, I was thinking about most mainstream religions. If you question the members of almost any religion, they will at first say they believe one thing, but on further questioning either back down or admit that they don't really know. Quite often theists will say something on the grounds that it 'might be true' even if they themselves think it highly unlike ...[text shortened]... ll say things like 'touch wood' - but nevertheless claim not to actually believe in such things.
Originally posted by moonbuswhat about those posting clear religious dogma in a science forum? Doesn't that qualify under the heading of troll? It's not like that hasn't happened. Could a computer do that?
Ah, yes. "Compartmentalization" is the technical term for that. Another example of that is astrology. A lot of people read the astrology column of the daily newspaper; if the 'prediction' comes true, they say they believe in astrology; if it doesn't come true, they still read the astrology column next day.
Originally posted by sonhouseI’ll lay my cards on the table here:
what about those posting clear religious dogma in a science forum? Doesn't that qualify under the heading of troll? It's not like that hasn't happened. Could a computer do that?
1. Always plays the same opening ‘gambit’ with insignificant variations of move order. Very machine-like.
2. Has no original thoughts. Everything he ‘knows’ is copy-pasted from the Internet. No ‘school book knowledge’ whatsoever. Very machine-like.
3. Never resigns, even when painted into a corner and in an obviously hopeless position. Very machine-like.
4. Shows no emotion. Shows no contrition when other’s feelings are hurt, for example. Seems not to comprehend when his own comments might be hurtful. Very machine-like.
5. Merely feigns taking wildly exaggerated offense at things to which no one else would. As if completely lacking sentience. (OK, might be Asberger syndrome.)
6. No sense of humor; fails to see irony in other’s or own remarks. Very machine-like.
7. Maintains theses wildly at odds with common sense. No apparent familiarity with basic laws of physical nature or traits of human nature. No ‘experiential knowledge’ whatsoever. Very machine-like.
8. Thorough-going literalism; no comprehension of metaphorical, allegorical, poetic, or other aesthetic dimensions of meaning.
9. Always opens with a sensationalistic topic guaranteed to elicit a response from humans, but cites sources which common sense tells you cannot possibly be reliable, roughly on the order of tabloid press articles claiming that so-and-so was abducted by space aliens and subjected to a rectal examination. What human would not respond to that? See, for example, the parallel thread on ‘giant humans’.
That’s all on one side. Maybe some of them are unimportant; I won’t argue that. But look at the number of them. On the other side all we’ve got is, only a human could be that obnoxious. But aren’t we merely projecting a motive into something where there might not be any motive at all? Like imputing anger to a lightning bolt or punishment to a plague?
In sum: we observe intelligence without sentience or common sense.
Conclusion: There is no such person as RJHinds. I submit that the persona known to us under that moniker fulfills all the criteria for AI. We’ve been Turing tested.
Open the pod bay door, Hal!
Open the pod bay door, Hal!
Open the pod bay door, Hal!
Originally posted by moonbusI don't think evidence that a given poster is cheating at chess (which has been pretty much proven in the past) is evidence that the posters forum comments are by an AI.
I’ll lay my cards on the table here:
1. Always plays the same opening ‘gambit’ with insignificant variations of move order. Very machine-like.
9. Always opens with a sensationalistic topic guaranteed to elicit a response from humans,
So basically trolling to some degree. I say that knowing what will elicit a response, takes human intelligence or a very sophisticated AI. It certainly indicates human rather than AI.
Originally posted by twhiteheadI agree with your first point: engine cheating at chess is not the same as being an engine. No inference from one to the other there.
I don't think evidence that a given poster is cheating at chess (which has been pretty much proven in the past) is evidence that the posters forum comments are by an AI.
[b]9. Always opens with a sensationalistic topic guaranteed to elicit a response from humans,
So basically trolling to some degree. I say that knowing what will elicit a response ...[text shortened]... akes human intelligence or a very sophisticated AI. It certainly indicates human rather than AI.[/b]
As for knowing what would likely elicit a human response (counter-trolling), it wouldn't take much programming to simulate that; after all, the google engine is a master at figuring out what people want to look at based on what they've looked at before. An SF algorithm could trawl sensationalistic web sites for search-phrases containing "Young Earth" or "evilution"--not hard to do-- and some forum member would be sure to take the 'bait'. Once the bait has been taken, the 'conversations' over at SF run in very predictable ruts.
1 edit
Originally posted by moonbusIf anything, I would expect someone attempting to get their program past the Turing test to either not play chess at all, or make sure the chess is of poorer quality or more human like.
I agree with your first point: engine cheating at chess is not the same as being an engine. No inference from one to the other there.
I do have to say that if the poster in question is an AI, it has passed the Turning test as far as I am concerned. It has completely failed a few other tests though including the 'sanity check' test.
Originally posted by twhiteheadYeah, I agree with that. The case is sol-ved, as Insp. Clouseau would say.
If anything, I would expect someone attempting to get their program past the Turing test to either not play chess at all, or make sure the chess is of poorer quality or more human like.
I do have to say that if the poster in question is an AI, it has passed the Turning test as far as I am concerned. It has completely failed a few other tests though including the 'sanity check' test.