Slashdot Mirror
The Profoundly Weird, Gender-Specific Roots of the Turing Test
malachiorion writes: Alan Turing never wrote about the Turing Test, that legendary measure of machine intelligence that researchers claimed to have passed last weekend. He proposed something much stranger — a contest between men and machines, to see who was better at pretending to be a woman. The details of the Imitation Game aren't secret, or even hard to find, and yet no one seems to reference it. This article explains why they should — in part because it's so odd, but also because it might be a better test for 'machines that think' than the chatbot-infested, seemingly useless Turing Test.
That would be a hard test. Most guys can not even pull it off. Lord knows I have seen a lot of guys online trying to pretend to be women who just can not even get the basics down.
Isn't this just another case of "The AI curse". As soon as a goal is reached, everyone declare that it wasn't "real AI after all", moving the goal post post-facto.
You're asking a machine to mimic something profoundly alien to it's nature, to put things on an equal footing the man should have to do so as well.
And please, let's not get in to the similarity/difference argument. Yes, the similarities between the minds of men and women far outweigh the differences, but the differences are still profound.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
What a minute, isn't this what IRC was invented for? For guys to go in to sex channels and pretend to be women and for IRC bots to pretend to be women too? I though this was already a settled matter...
Comment removed based on user account deletion
the headline might as well have been, "one weird trick to designing a turing test"
The Imitation Game is given with an example of two human participants where a third has to interrogate both and decide which is the real woman, computers aren't mentioned in the example (The paper is linked by wikipedia so it is indeed easy to find). The test involving computers places the computer as pretender against a man (the real human) - so no the computer does not have to pretend to be female, its just given to explain the test.
Also the turing test was not beaten - apparently the guy responsible for it is also known as mister cyborg and even claimed that the "winner" used a super computer (it ran on an underpowered laptop).
to understand how a machine could be made to pass the Turing test (or the woman test) honestly and thoroughly. To do so, it would have to understand arbitrary human statements and questions: not just "why is the sky blue?" (relatively easy) but "why doesn't my wife understand me?" and "is the real rational, as Hegel posited, and if so (or not) why (not)?" Note that the machine could reasonably pretend to know nothing about Hegel, but it would have to react like a normal human being. No obfuscation such as pretending to be foreign, a child, thoroughly ignorant, or befuddled by drugs should be accepted.
Going a little further, it would have to cope with (very) simple jokes such as "I asked my dog which team would win the World Cup"/"What did he say?"/"Nothing. HE'S A DOG".
I am sure that there are many other solipsists out there.
May have been an expert at pretending to be a woman.
Of course, not in the league of J. Edgar Hoover...
"Flyin' in just a sweet place,
Never been known to fail..."
Maybe the key to the Turing test is one question -- Top or Bottom?
"Passing" the Turing Test is about as valid as Asimov's Three Laws of Robotics.
Proverbs 21:19
not quite, fooling someone into believing the opposite. Catchers playing pitcher, or bananas to someone's doughnut (japanese phrase there).
Just read Turing's fucking paper linked in the article. It is clearly about a machine competing with a human to convince a third person that he is the human and that the human is the machine. The interlocutor is promised that he is conversing with both a machine and a human, but does not know which is which. The man vs woman game is just an initial example to make it clear what the game is about. Then Turing goes on to replace the man by a machine, but also implicitly assuming that the interlocutor knows that he now has to decide between a machine and a woman. As for the human side, it doesn't matter whether it's a woman or a man.
So it seems that Turing's test was not won by man or machine, but a synthesis of the two: man + internet
If a woman programs the machine, then this version of the Turing test should be no less difficult than the accepted version.
excitingthingstodo.blogspot.com
Rachael: The egg hatched...
Deckard: Yeah...
Rachael: ...and a hundred baby spiders came out... and they ate her.
Turing was gay, as such did he have some culturally "feminine" interests or ways of thinking, or was he more a "man-gay"
Not so weird. There may be plenty of sociological reasons one could consider for making this gender (or sex) specific that are probably pretty valid. (Men and women have dramatically different ways of understanding communication, and this was not unknown in the 1950s.)
But if the rest of us are forced to consider Turing's homosexuality as a person indivisible from his work as a visionary (OMG he was a thinker, and he was gay!), it's only fair to consider that trait in the analysis of the work he did...
Hire a Linux system administrator, systems engineer,
Alan Turing never wrote about the Turing Test
Didn't he? Wikipedia has this to say:
he proposed a third [version] in 1952. In this version, which Turing discussed in a BBC radio broadcast, a jury asks questions of a computer and the role of the computer is to make a significant proportion of the jury believe that it is really a man.
That sounds like the usual formulation of what get called Turing Tests these days..
One of the people quoted in the article seems a bit confused as well:
“Turing never proposed a test in which a computer pretends to be human,” says Karl MacDorman, an associate professor of human-computer interaction at Indiana University. “Turing proposed an imitation game in which a man and a computer compete in pretending to be a woman."
from which I can only infer that Karl MacDorman doesn't consider women to be human.
systemd is Roko's Basilisk.
So, what about switches? Surely it's not an "all one or the other" trait.
Lost at C:>. Found at C.
According to the article:
Turing never proposed a test in which a computer pretends to be human,” says Karl MacDorman, an associate professor of human-computer interaction at Indiana University. “Turing proposed an imitation game in which a man and a computer compete in pretending to be a woman.
What the heck is "shingled" hair? Do I go to a barber, a hairdresser, or a roofer?
Gamingmuseum.com: Give your 3D accelerator a rest.
He wanted a sexbot that he couldn't tell was a robot. Classic Turing.
You don't get "profoundly weird" and "gender specific" being used to describe something that is also "better" too often. Might be a reason for that... maybe "better" isn't the right word for it.
"The details of the Imitation Game aren't secret, or even hard to find, and yet no one seems to reference it."
Except, well, at least four of the stories I've seen on the Turing test this week. It really doesn't seem that obscure.
Women don't talk about poetry, math, and chess?
This alternate test sounds really easy:
Q: Add 34957 to 70764.
ELIZA: Don't you ever say Hello?
Man: What does this have to do with being a woman?
girls are weird.
profoundly weird.
Yes, but especially in Turing's time men and women would typically have rather different perspectives on such things thanks to very different typical roles in society (And how about the scandalous Spanish rebranding of the Vizier in chess to the Queen which took Europe by storm, making the most powerful piece on the board female. Though admittedly that was well before Turing's time). By forcing men to try to bridge that social gap in their conversation he was putting them at a disadvantage not totally unlike the gap between a human mind and one built on a foundation of rigorous logic. At a minimum, at least both contestants were pretending to be something they were not.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
that's merely another possibility for a test, can a switch-hitter fool someone into thinking they're not
isn't that creepy.
- Zav - Imagine a Beowulf cluster of insensitive clods...
Didn't Chris Hansen of Dateline perfect this? He's made an entire career out of impersonating underage girls to get guys to show up some place so he can film them in his "To Catch a Predator" series.
Anyone holding a Turing Test isn't really holding a true Turing Test. Turing didn't define a duration but it's pretty obvious that five minutes isn't long enough to hold a conversation, it's only long enough to ask a series of questions.
To hold an actual conversation you'd need an hour, maybe more. To last that long the bot would need to learn things about the interrogator just like we learn things when we talk to eachother. A bot that could do that could obviously do a lot more as well.
That's not to say the current Turing Test isn't useful. A successful bot could be applied to customer support systems and give insights into language. But it's misleading to indicate that the current incarnation of the Turing Test is testing what Turing intended.
I stole this Sig
Research has long proven that women have a different corpus callosum than men (which makes them better information synthesizers), where their splenium has more parallel channels of communication between the hemispheres of the brain (10 to 15 channels, whereas men only have 1 or 2 channels). Swedish researchers have proven that gay men have female-typed corpus callosum's. Turing's Imitation Game could have originated because as a gay man he realized he was better at pretending to be a woman than other men were...
-=/\- Jizzbug -/\=-
So Judith Butler quite famously put forward the thesis that gender in society is primarily a matter of social role performance, as opposed to any kind of physiological reduction. Being Man or Woman (as opposed to Male or Female, though Butler does also throw those conceptions into question) in a particular social setting is not so much about how someone is socialized or their brain structures as such, but more how they subsequently go on to express certain behaviours and phrases, modes of self presentation or verbal habits that mark them out in the social sphere as being of one particular category or the other. Moreover, that this kind of performance type view helps us explore exactly what it is about the gender roles of, say for example our society, that gives rise to both traditional, heterosexual patriarchal views about men and women and also new kinds of gender expression that might subvert those old standards or work towards values that we want to encourage in the world we live in.
One thing Butler often talks about is how tied up sexuality is in issues of gender performance. After all, performing gender in a culture of rigidly enforced masculine heterosexuality is a very different challenge to performing gender in a culture where gay men will not be (as) institutionally judged as deviant from permissable social standards. Alan Turing would have spent much of his life not in a position of denial about his gender but rather in a position of having to practice his masculine gender outwardly in such a way as to deliberately deflect suspicion from him as a gay man. That is, in affirming his gender, the world he lived in would have proscribed him to do so in a very self-negating way.
It's really interesting to see that Turing might have looked at gender in similar lines to the way he looked at intelligence - as something that needed to be outwardly demonstrated in order to be put forward for effective analysis. It speaks to something of the struggle he must have faced in his private life. Huge respect for him for speaking up about it!
Myu:
Turing's imitation game was a thought experiment. To explain it simply to an audience unfamiliar with the idea, he started with man against woman, then proceded to man against machine. The specific genders were not important or significant in the thought experiment, just the existence of some difference between the contestants that could potentially be spoofed over a teletype. There was nothing gender-specific or weird about it.
Also, either the author of the article has a listening comprehension problem or the assitant professor quoted in the article has a reading comprehension problem.
Look at Turing's original article. It says that the imitation game is played between a man (A), a woman (B), and a player C. C has to decide among A and B who is a man and who is a woman. Now, the _man_ is replaced is a computer and we ask if C will perform as well or poorly as before.
So in Turing's version we have a computer A pretending to be a woman to C, and a woman trying to convince C that she is the woman.
Turning's original test _does not_ have a man and a computer pretending to be a woman to a judge.
Q: What do you want for dinner?
A: Anything's fine.
Q: How about X?
A: No.
Q: How about Y?
A: No.
Q: How about Z?
A: No.
Dear Slashdot: next time you want to mess with the site, add a rich-text editor for comments.
imagine the turmoil in online games and MMOs(especially those with rp or erp communities) and chat channels if one could reliably detect the difference between a guy and a guy acting the part of a girl?
As mentioned above, Turing never proposed to have a computer attempt to imitate a woman. What would be the point? Why would imitating a woman be harder than imitating a man? The interpretation discussed in the article is not "debatable". It is evidently false when you actually read the paper.
There is a slight ambiguity at the end of page 1 of the paper, where Turing writes that the man of the imitation game (the one pretending to be a woman) could be replaced by a machine. Does it mean that the machine will try to imitate a woman, or that the machine will "merely" imitate any human being of its choice? Turing never mentions gender roles again in the paper; and, exactly one page after introducing the imitation game for machines (that would be page 2 overall), Turing settles the ambiguity once and for all by writing this:
" (...) It might be urged that when playing the "imitation game" the best strategy for the machine may possibly be something other than imitation of the behaviour of a man. This may be, but I think it is unlikely that there is any great effect of this kind. In any case there is no intention to investigate here the theory of the game, and it will be assumed that the best strategy is to try to provide answers that would naturally be given by a man."
The original paper written by Turing (titled "Computing Machinery and Intelligence"), in which he describes the imitation game, is readily accessible from multiple sources on the Internet. Check for yourselves. And I suggest you read it to the end, the paper is also a very easy introduction to Turing Machines, multiple realizability, and other cool stuff, all of it explained with incredible clarity by one of the most brilliant men in recent history. That will certainly do you more good than to read "popsci.com" articles by authors who can't be bothered to read more than a single page about their subject matter before they start writing nonsense.
PS: Note however that Turing also wrote: "In the process of trying to imitate an **adult** human mind we are bound to think a good deal about the process which has brought it to the state that it is in." Take that, Eugene Goostman.
I am a pretty woman and I hope everybody believes me.