Russian Chatbot Passes Turing Test (Sort of) 236
CurtMonash writes "According to Ina Fried, a chatbot is making the rounds that successfully emulates an easily-laid woman. As such, it dupes lonely Russian males into divulging personal and financial details at a rate of one every three minutes. All jokes aside — and a lot of them come quickly to mind — that sure sounds like the Turing Test to me.
Of course, there are caveats. Reports of scary internet security threats are commonly overblown. There are some pretty obvious ways the chatbot could be designed to lessen its AI challenge by seeking to direct the conversation. And finally, while we are told the bot has fooled a few victims, we don't know its overall success rate at fooling the involuntary Turing "judges.""
How long do we have ... (Score:2)
( http://youtube.com/watch?v=yomx7bXMf2U [youtube.com] )
5 years ? I doubt it.
Re:Who Loves You, Baby? Putin Loves You, Baby !! (Score:5, Interesting)
Re: (Score:2)
Re: (Score:2)
All well and good... (Score:5, Funny)
In fact, the chat bot side of things is wholly superfluous to what I want if I'm being honest.
What I really need... (Score:3, Funny)
Then I can run the bot, play Crysis and just show up at the right place and time on Saturday night.
Bonus points if convinced women are attractive.
Re:What I really need... (Score:5, Funny)
Re: (Score:2)
In Soviet Russia (Score:5, Funny)
Re:In Soviet Russia (Score:5, Funny)
Re:In Soviet Russia (Score:5, Funny)
Jubii had such a robot (Score:5, Informative)
Getting financial details is probably new, but that was predictable.
Re: (Score:2)
And how much more appropriate that would have been.
Re:Jubii had such a robot (Score:5, Interesting)
On top of that, there is the whole chat medium. Anyone who has ever done a lot of IM/IRC/whatever knows that it's not uncommon to type the wrong thing in the wrong window/channel, so the occasional out of nowhere sentence that would never pass in a one-on-one environment, will pass there because the signal to noise ratio is lower.
Still, I'd be interested to see the code, and see how well it deals with non sequiturs.
Re:Jubii had such a robot (Score:5, Funny)
I enjoy rhubarb.
Re: (Score:2)
Re: (Score:3, Interesting)
Somehow I find that idea even more disturbing.
Bull (Score:4, Funny)
Re: (Score:2)
Old News... (Score:5, Funny)
Re: (Score:2)
Re:Old News... (Score:5, Funny)
Re: (Score:2)
English lyrics (Score:2)
I know a bot,
her name is Anna, Anna is her name
And she can ban you, ban you so hard
She cleans up our channel
I want to tell you that I know a Bot
I know a bot,
her name is Anna, Anna is her name
And she can ban you, ban you so hard
She cleans up our channel
I want to tell you that I know a Bot
That always watches everyone in our channel
And sees that there is no trouble in here
And it can no way be taken over
And remember, I know this
Re: (Score:2)
And therein lies the fun part. (Score:2, Insightful)
Re: (Score:2)
My point is proven yet again, that the vast majority of humanity lacks the simple survival skills that would make us worthy of propagating and passing on our genes...
Chatbots aside, looking for sex with females, however cheap and easy, seems like it has historically been an effective way of propagating and passing on genes.
Oh, sorry, "worthiness" of reproduction is perhaps a separate matter from effectiveness - that's a matter of eugenics, really. Perhaps you would be happier if these chatbot-seeking individuals were to worshipfully obey some authority that tells them they are unfit to reproduce?
Re: (Score:2)
Re: (Score:2)
Nowadays and in some countries, sure. Most of history and the world, it's been otherwise and they did just fine. Mail-order brides worked out OK too. Your premise is shaky.
Re:Correction. (Score:5, Insightful)
I would hate to live in your world. I value compassion as highly as intelligence, and letting somebody suffer and die while you are able to help, on some aloof philosophical or eugenic basis is simply inhumane. Is there anybody in your life that you care about? Would you try to help them if they were suffering? Would you care for them less if they'd made a mistake and their suffering was in some way self-inflicted?
You should value people based on what they are or what they can give, not on what they are not or what they lack.
Re: (Score:2)
My point is proven yet again, that the vast majority of humanity lacks the simple survival skills that would make us worthy of propagating and passing on our genes... evolving and surviving, if you would.
I'm afraid the evidence points to the contrary. As a species we don't seem to have any problems surviving.
Moreover the extraordinary success of the human species is mostly, or maybe entirely due to its incredible survival skills. These skills are also very widely adaptable, so we can thrive in whatever environment we find ourselves in. We are not limited by the environment in which we evolved (which certainly didn't include bots on irc) and as soon as a new threat is identfied, we quickly spread the
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
I can use my life (once more) as a case in point.
I've learned at least 6 or 7 different trades, mastered several skills, speak several languages fluently (read, write, etc), have traveled some of Europe and some of the USA, and have gotten to the point where I just plain don't give a shit. AND ALL THAT before I hit 30... Why? What's the point? To have my name on a plaque? What does that accomplish? What do **I** get out of it? I'm not a damn cell. You might be, but I'm n
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
the vast majority of humanity lacks the simple survival skills ... ready obedience ....
Obedience to accepted authority IS a survival skill. If a tribe of hunter-gatherers (or worse, farmers) argues every decision, they starve.
Your larger point -- fools deserve to be parted from their money -- is economically true. These people are idiots -- but the human race needs idiots, because sometimes figuratively watching grass grow really is important for the survival of the tribe.
Re: (Score:2)
Of course, the vast majority of humanity I've met quickly reminds me that in fact, homo sapiens IS just a stupid monkey, of which a few mutated to great intelligence and managed to drag the rest of the unthinking herd (for they truly are a herd, not a pack of hunters) out of the primordial slime.
That bein
Re: (Score:2)
WTF? This is not even a Turing test. (Score:5, Informative)
Re: Restricted Turing Tests (Score:5, Informative)
One of the reasons that AI researchers moved away from the pure test is that it becomes more about "gaming the conversation" than a test in real intelligence.
People have no trouble "abusing" the conversant if it is part of a test with a bot. Therefore, the *person* also gets subjected to degenerate forms of conversation until he/she "authenticates as a person".
(Really, someone just needs to put a few million of funding into some defensive conversation routines to make their perceived performance go through the roof. The problem so far has been everyone duplicating everyone else's efforts.)
Although I have done thought studies of the reduced level of "intelligence" in chat rooms to begin with, they don't feature the same "bust the knowledge domain" questions seen in typical Turing contests. In fact, asking those questions earns you *ridicule* in other chat environments.
Therefore, by "disallowing" the artificial questions, if the chatter failed to detect the BotHood of the conversant on the other side side by side with real people, it passes a form of Restricted Turing.
Re: (Score:2)
But if people assume that they are dealing with a real human (and it doesn't even enter their mind that the person on the other end MIGHT be a computer), then I don't know how much credence you can put in the results.
Turing probably was not serious about this test (Score:5, Insightful)
Turing was a mathematician, which came through in all his thinking, including devising the Turing Test. When faced with questions like "can a machine ever be intellignt?" it is virtually impossible to answer this directly because, firstly, how do you define intelligence; and secondly,how do you measure intelligence?
Mathematicians **hate** imprecise questions because they cannot be proven or answered satisfactorily.
When faced with this problem, Turing used the well loved mathematical method of reductio ad absurdum: if you cannot tell the difference between a human and a machine, then it is absurd to claim the human is intelligent but the machine is not. That neatly sidesteps all the impossible to answer questions like the precise definition of intelligence. Typical mathematician wriggle out move.
Is the Turing Test practical? Well perhaps not. Machine intelligence (whatever that means) can be useful without the machine holding a conversation with you. Annoyingly it has soaked up a lot of effort with people building talkbots instead of getting on with more practical aspects of machine intelligence.
Re: (Score:2)
It's not a wriggle-out move. It's logic. (Score:2)
> the human is intelligent but the machine is not. That neatly sidesteps all the impossible
> to answer questions like the precise definition of intelligence.
> Typical mathematician wriggle out move.
No it's not. It is a necessary law of logic. If two entities are not distinguishable, they must be the same entity, or contradictions may arise.
Re:WTF? This is not even a Turing test. (Score:5, Funny)
Re: (Score:2)
I think in this case, the men see this thing offer to chat about sex and their brain goes out the window which is why they don't notice at that point. I mean hell, given all of the bad typing and spelling and inability to correct typos I see out there, even if this thing talks in broken Russian, they probably think the girl is just blonde
Re:WTF? This is not even a Turing test. (Score:5, Insightful)
The key part of the turing test, to me, is that the judge must know they are engaged in the test. The best example of this is Eliza (read about it [wikipedia.org]). To someone critically examining it, it does not past the turing test. To someone expecting a therapist, most of its responses do make sense. The point is that if you're not trying to trip up the chatbot it's not hard to fool someone.
Re: (Score:2)
Re: (Score:2)
Link? Closest I could find is this http://www.sciam.com/article.cfm?id=the-truth-about-online-da [sciam.com] written by Dr. Epstein. But that has nothing to do with being fooled by a chatbot. It would be quite shocking if he was fooled seeing as he is a psychologist and he designs AI tests.
This is nothing new. AI researchers were fo
Re: Turing Test Family (Score:2)
Restate the problem this way:
"Is this really a hot babe looking for action, or is it (something) trying to scam me?"
People trying to be amorous and hook up qucikly have reduced the converasation domain. Then when some really weird answers come back, you do start trying to figure out "which agenda" is going on. I see little difference between a Bot scammer and a foreign scammer; both would use weird phraseology.
The ever-rising bar on true AI (Score:4, Insightful)
Now we have chatbot that can fool some people some of the time, so the bar has been raised on "true AI" to say that computers can't fool expert suspicious Turing test judges. This too will fall. Human intelligence is very slowly growing (they actually reset IQ tests every decade or so) but computer intelligence is growing much much faster.
Re: (Score:2)
The real question is whether the Turing test is an actual valid test of AI. If a simply programmed chatbot on a relatively average computer can pass it, then that's pretty good evidence that the Turing test isn't testing for actual "intelligence".
Re: (Score:2)
(Playing the advocatus diaboli here)
Re: (Score:2)
Re: (Score:2)
How do you determine that the software doesn't understand what it's doing?
What if the software was actually a complete simulation of the physics of the human brain? Unless you believe in a non-physical component of the mind, then the si
Re: (Score:2)
Re:The ever-rising bar on true AI (Score:5, Insightful)
Indeed, it's easy to show that fooling some people some of the time doesn't require anything even approaching AI. Consider a bot that simply repeats a set of ten sentences in a fixed order: if those sentences were chosen well enough, then some people might easily believe that they were having a real conversation. But I really don't think you'd argue that a bot that simply repeated a set of ten sentences in a fixed order displays any sort of intelligence, no matter how many unsuspecting people happen, by random chance, to feed it lines that cause its responses to look relevant.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Yet still people argue that it does not really count.
It doesn't.
This is analogous to the "computers can never beat people at chess" meme
No it isn't. Saying "that can never happen" is completely different from saying "it hasn't happened yet".
the bar has been raised on "true AI" to say that computers can't fool expert suspicious Turing test judges
For the Turing test, the bar has always been that high. It didn't get moved up there recently. (BTW, the term you are looking for is "strong AI [wikipedia.org]".)
This too will fall.
Probably, assuming nothing crazy happens like a vacuum metastability event. But it's immature to claim we are anywhere close.
Re: (Score:2)
Now we have chatbot that can fool some people some of the time, so the bar has been raised on "true AI" to say that computers can't fool expert suspicious Turing test judges. This too will fall. Human intelligence is very slowly growing (they actually reset IQ tests every decade or so) but computer intelligence is growing much much faster.
While it's true that both are rising, I don't think the comparison with chess is valid. If you place the same chess engine on a 3GHz machine instead of a 300MHz machine, we know it will be better and you can quantify how much too. A chatbot on the other hand may not, unless you can find more meaningful work for it to do, it can't just check a "conversation tree" to greater depth. Three of the things I've found most lacking is implied states, states not specified and identifying non-sensical statements.
As a
Re: (Score:2)
Re: (Score:3, Informative)
The "no true AI" paradox (Score:2)
Eliza says- (Score:5, Funny)
Re: (Score:2, Interesting)
"Nice weather."
"Yes, nice weather."
"It might rain this afternoon."
"Rain? You think so?"
"You're elizing again!"
Turing test extra credit (Score:3, Funny)
http://xkcd.com/329/ [xkcd.com]
Obligatory Futurama reference (Score:5, Funny)
Fembot: It's true. I disguised myself as a easily-laid woman so I could rule the Russians.
Russian guy: But why?
Fembot: Why? Why? I came here from a faraway planet. A planet ruled by a chauvinistic Manputer that was really a Manbot. Have you any idea how it feels to be a Fembot living in a Manbot's Manputer's world?
This test is very easy (Score:5, Interesting)
A decade ago I wrote a perl script for sirc that had 40 sentences and would just reply one picked at random (uniformly) every time it would get a private message. Hence it was not taking into account neither what was the message it just received to it (a la Eliza) nor what it had said before. It was not even waiting before replying, hence would type the respones in a tenth of a second.
It happened several times that people would talk with it for more than an hour. If I remember correctly the record was 1h45min ...
For the Turing test, the tester has a strong prior that the testee may be a computer. This is not the case here, and the prior for this to happen is so low that it's impossible for a layman to come with that explanation. What happens is that people think inconsistencies in the speech of their interlocutor is due to technical problems (sending message to the wrong person, lag, complexity of the program the person use, etc.)
Re: (Score:2, Insightful)
It works both ways
Re:This test is very easy (Score:4, Interesting)
Seems like (legal) vaporware (Score:5, Informative)
Re: (Score:2)
There, fixed that for you. Seriously, any woman that does run a mile and/or get a restraining order after being told you got her phone number and photos off the Internet probably is a Russian Fembot (with machine gun jubblies [ew.com].)
They're chatbots?! (Score:4, Funny)
Emacs Girl? (Score:3, Informative)
Hello?
How do you do? What brings you to see me?
I am a hot russian girl
Why do you say you are a hot russian girl?
Are you looking for a girlfriend?
What do you think?
I'm a double D and love to have fun
fun what ?
What makes you believe and love to have what?
Aren't you a curious one...
Earlier you said and love to have what?
Can I have you're phone number??
Maybe
well duh! (Score:2, Funny)
Humans failing Turing Test. (Score:2)
Turing test involves comparison (Score:2)
If the judge was communicating with just one, there probably would be many cases where real people wouldn't pass. But when you're having a side-by-side comparison between humans and AI, it's currently very easy to tell which is which.
Possible Good Thing?? (Score:2)
The scammed (not so bright) 'victim' loses identity, credit, etc... and becomes far less desirable as a mate; therefore having less of chance to reproduce.
What about ... (Score:2)
I've seen quite a few posts/articles/etc. on various systems on assorted subjects where the originator of the thread submits some standard dogma about Jesus Christ/Muhammad/whomever and either never respond to subsequent queries or respond with some obscene vitriol about how questioning faith is the ultimate blasphemy.
Heck, I could knock one of these 'bots together in a few hours with Perl.
the problem is the user (Score:3, Insightful)
That's not a good test for AI. Research shows that men go crazy while talking with beautiful women. So, sexuality temporarily shuts down their intelligence. You can't test for AI while employing sexuality.
Chris Hansen (Score:2)
Just want we wanted (Score:2)
Re: (Score:2)
Jenny18 (Score:4, Informative)
http://virt.vgmix.com/jenny18/ [vgmix.com]
May be old news, perhaps even fake (Score:2)
The CyberLover [cyberlover.ru] program site doesn't do much. None of the links work, including the one for sample chat logs. The site says "Copyright 2005-2006", so this has been up for a while. The site was trying to recruit "affiliates", for a program that sells for only $4.95. This looks like an idea that didn't work.
It's actually not as hard as you think. (Score:5, Interesting)
To plant him, we simply made a free page on some blog with some personal details and put his IM up there and waited to see what happened.
We eventually shut him down because people were becoming way too personal with him. One girl had an ongoing series of conversations with him about how she was recently raped. His mouth became rather foul when my roommate decided to have him initiate a conversation (he had a whitelist of known 'admin' screen names who could then order him to say something specific to a specific screen name) with screen names linked to hate groups. Another guy just wanted to convert him to evangelical Christian. It was way too simple to write a bot to make many, many people think is real. Some people did figure it out, so if someone ever brought up 'bot' in a conversation they were immediately added to a blacklist so as not to corrupt the conversation database.
The biggest giveaways? "u type too fast" (we eventually added a delay to solve that issue) and "u only type something when I do" (by this time I had already decided it was time to shut down the bot for good). It was a lot of fun until he started hurting people... if I ever resurrect him he will have a pre-set kill limit.
~Ben
Re: (Score:2)
How'd it handle slang? Could it 'andle droppin' lett'rs? Would it agree to have a lovely bunch of coconuts, all standing in a row? Would it parse/react to "Hair 3.14159"?
There are a lot of ways to find a bot... but as pointed out before, if you ALREADY believe, you're screwed.
Re: (Score:3, Funny)
I had similar issues with a bot I set up. I put it in an IRC chat room and it was great fun until it started repeating nasty insults it had overheard.
I deleted the database and fed it some other text to learn from. Interestingly, if you feed a chatbot the scripts of the Star Wars trilogy, it spews random nonsense whenever it types anything.
Re: (Score:2)
You'd see people respond the same way to books.
But that's because these ARE the few strange people who respond that way. The test doesn't really mean much, it's targeting way too few people.
What is remarkable is that it might be quite profitable.
Intention (Score:4, Interesting)
Phone Sex Bots in Bruce Sterling Story (Score:2)
"The software just picks words at random out of the customer's own sick, pathetic rant! Whenever he stops for breath, it feeds a question back to him, using his own vocabulary
-kgj
As Q would say (Score:5, Funny)
Turing Test is Bogus (Score:2)
That's total BS. Of course it depends on the human. I suppose there's some kind of NP-complete version which says any AI tested OK by Marvin Minsky is intelligent enough, though Minsky might just be playing favorites. Minsky might not pass many humans on the test, who are the kind who fall for this bot.
The fact is that the whole idea tha
Permutation City (Score:2)
Re: (Score:2)
No, you can't pass the Turing test by fooling stupid people with a computer program. That problem is trivially solved. The goal on Turing test is to fool the judge, who knows that there is a 50% chance that you are a computer program. It is much harder.
Extra bonus is given if after the test, Judge no longer knows whether he/she is a human or not.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)