[Dragon]

You don't get it, do you? Were you really so lucky to grow up only among equally or more intelligent people? Probably easier in whatever civilized place you live. Well, I had to grow up among a band of idiots. The only way not to become like people you live with is to distance yourself from them. As such, I had a choice of being an idiot obsessed with sex and alcohol, or resolving to put myself above all this. One can not function in a large group without taking up it's characteristics. Cutting them off wasn't so hard, really, they were so low that even you would've felt superior should you meet one of them (unlikely, unless you take a trip to Poland). It's not a superiority complex, it's just fact. I have to say, until I actually saw those people, I was, just like you, naively thinking that all people are equal and all that. Actual contact with people set me straight quickly, and you know what's funny?. This approach worked, and I'm more today than any one of those idiots could ever hope to be. It wasn't by chance (nor money - schools in Poland are free) that I went to one of two best high schools in my city, and I was one of the 3 people from that school in there, and the only one from my class.

Besides, that doesn't matter. You're ignoring a point I was making, and derailing the thread. If you don't have anything to say on topic of the AI, don't say nothing. I won't say anything more, either. I invite you to try conversing with rural 13-year olds from some backwards country one day. You'll either see my point, or find yourself right at home with them... Either way works for me.

Back on topic, is that chatbot available online? The article quoted provided a link, but it doesn't work for me.

[General Battuta]

believe it or not other people were bullied at school too, some of us just dealt with it through personal growth and maturity rather than retreating into childish superiority complexes

I would have said this but less harshly. Sounds like you had a tough childhood, but not an unusual one.

[karajorma]

Can we get back on the subject please?

[Dragon]

Sure. But there isn't much more to be said about the chatbot. It's not particularly smart, could probably pass for a somewhat moronic human for a while, and doesn't do anything revolutionary.

I formulated an argument that was ignored in favor of personal discussion by PH, so let's try again. Turning test isn't the best way of determining a computer's intelligence. In fact, I'd say that there can be no definite intelligence test one could use to compare computers to humans, because human intelligence varies greatly. Even the most advanced chatbots do not really understand what they're saying, nor can they create anything new, or form an independent opinion on a subject. IMO, a true, strong AI would need to be able to do that to be compared to humans. Anything else is just a souped up database search engine with a natural language interpreter. Those two would be important parts of any AI (our memory and reading ability, respectively), but there's more to true intelligence than that.

I would have said this but less harshly. Sounds like you had a tough childhood, but not an unusual one.

You're right, of course. I was unusual, not what happened to me. I react differently from most, and I'm prone to making observations and pondering the world I live in. Indeed, people I discussed my worldview with agreed with me. They didn't came to those conclusions themselves because they didn't think about it, nor observed morons so long and so closely (I had a choice between morons and wheat. Wheat is less annoying, but also less entertaining). I wouldn't be so confident in what I'm saying if I didn't discuss it before. My experiences are relevant only in that I've seen a lot of stupidity, and thus can probably identify and classify it.

[Scotty]

I would definitely appreciate that whole subthread coming to an end.  This is a topic on computers, not your past experiences.  Karajorma has already made that clear.  Next time it comes back up should not happen.  If it does, there will be problems.

[Dragon]

Fine, I just hate leaving things just... dangling like that. Besides, we've already agree that the bot doesn't really do what was claimed. There's hardly anything but random banter left to say. And derailing something is easiest at the end of the tracks, so to speak.

[karajorma]

If you have nothing to say, stop speaking. Final warning.

Scott Aaronson rips this bull**** Turing Test apart in about 2 questions.

Glorious, glorious takedown indeed.

I was glad to see the addendum that Veselov isn't responsible for this nonsense and has no illusions that his program has passed the Turing Test. When even the guy who wrote the program thinks it's being overhyped that's a pretty strong sign that people have gone a little bit off the deep end about it.

[Phantom Hoover]

Getting back to the main topic, the reason the Turing test is formulated the way it is stems from Turing's own theoretical work on abstract computing machines. Two Turing machines are considered equivalent if they always produce the same output given the same input; and since thanks to the halting problem you can't meaningfully compute any non-trivial information about a TM that's the only way to really think about it. Applying this concept to the question of whether a machine can be as intelligent as a human brain and taking some artistic licence you arrive at the idea that interrogating the computer to see if it can be discerned from a human is an important first hurdle along the way.

[Luis Dias]

While I never bought the hype from the very moment I saw the headline, I must say that Aaronson is either being a jackass or a fool. Yes, of course his conversation with the internet bot he found is abhorrently godawful, but this has been the case with all internet bots, whose more complex versions powered up by potent machines dedicated to a single conversation in a more direct talk (in bot contests, etc.) are waaaay more believable and interesting than that conversation implies. The test was not made with the "kind" of mind Aaronson talked to, in a nutshell. Or, if the main engine was more or less the same, the amount of resources the net bot had at its disposal is obviously considerably smaller in all sorts than the original that went through the tests.

I would very much like to read the original scripts of the talks that alledgedly fooled the judges.

[Ghostavo]

[karajorma]

I would very much like to read the original scripts of the talks that alledgedly fooled the judges.

As would I.

Who wants to bet they are never made available, let alone published in a peer reviewed journal then? People seem to be forgetting this is Captain Cyborg that we're dealing with. 

[Ghostavo]

http://www.theguardian.com/technology/2014/jun/09/eugene-person-human-computer-robot-chat-turing-test

It's kind of painful to read some of the exchanges.

[Luis Dias]

You think? Could you tell me what are you? I mean your profession.

Oh what a fruitful conversation! :-)

[Sololop]

Yeah, i've seen a lot of the examples. It's pretty pathetic, I'm certain I've seen better chatbots in the past? Either way, it really shouldn't be able to fool anybody. It repeats itself often and acts very strange...

[pecenipicek]

i've personally spent more time than is healthy trolling iGod... so yeah...



if nothing else, it was a sorta "hur durrr" way to pass a slow afternoon...

[Klaustrophobia]

I've personally interacted with with a bot that I probably wouldn't have known was a bot if I wasn't told.  Granted, I wasn't trying to figure out if it was or not.  It was a virtual teacher designed to help with basic programming.  It also didn't speak very much because I in no way needed help with the trivial programming task that was the test case, but the few times it did it was quite impressive.  It even asked me if I learned to code in C because it thought my style was similar.