Turing the mind - PeachParts Mercedes-Benz Forum

Botnst · #1 02-16-2006, 08:29 AM

The Trouble with the Turing Test

Mark Halpern

n the October 1950 issue of the British quarterly Mind, Alan Turing published a 28-page paper titled “Computing Machinery and Intelligence.” It was recognized almost instantly as a landmark. In 1956, less than six years after its publication in a small periodical read almost exclusively by academic philosophers, it was reprinted in The World of Mathematics, an anthology of writings on the classic problems and themes of mathematics and logic, most of them written by the greatest mathematicians and logicians of all time. (In an act that presaged much of the confusion that followed regarding what Turing really said, James Newman, editor of the anthology, silently re-titled the paper “Can a Machine Think?”) Since then, it has become one of the most reprinted, cited, quoted, misquoted, paraphrased, alluded to, and generally referenced philosophical papers ever published. It has influenced a wide range of intellectual disciplines—artificial intelligence (AI), robotics, epistemology, philosophy of mind—and helped shape public understanding, such as it is, of the limits and possibilities of non-human, man-made, artificial “intelligence.”

Turing’s paper claimed that suitably programmed digital computers would be generally accepted as thinking by around the year 2000, achieving that status by successfully responding to human questions in a human-like way. In preparing his readers to accept this idea, he explained what a digital computer is, presenting it as a special case of the “discrete state machine”; he offered a capsule explanation of what “programming” such a machine means; and he refuted—at least to his own satisfaction—nine arguments against his thesis that such a machine could be said to think. (All this groundwork was needed in 1950, when few people had even heard of computers.) But these sections of his paper are not what has made it so historically significant. The part that has seized our imagination, to the point where thousands who have never seen the paper nevertheless clearly remember it, is Turing’s proposed test for determining whether a computer is thinking—an experiment he calls the Imitation Game, but which is now known as the Turing Test.

The Test calls for an interrogator to question a hidden entity, which is either a computer or another human being. The questioner must then decide, based solely on the hidden entity’s answers, whether he had been interrogating a man or a machine. If the interrogator cannot distinguish computers from humans any better than he can distinguish, say, men from women by the same means of interrogation, then we have no good reason to deny that the computer that deceived him was thinking. And the only way a computer could imitate a human being that successfully, Turing implies, would be to actually think like a human being.

Turing’s thought experiment was simple and powerful, but problematic from the start. Turing does not argue for the premise that the ability to convince an unspecified number of observers, of unspecified qualifications, for some unspecified length of time, and on an unspecified number of occasions, would justify the conclusion that the computer was thinking—he simply asserts it. Some of his defenders have tried to supply the underpinning that Turing himself apparently thought unnecessary by arguing that the Test merely asks us to judge the unseen entity in the same way we regularly judge our fellow humans: if they answer our questions in a reasonable way, we say they’re thinking. Why not apply the same criterion to other, non-human entities that might also think?

But this defense fails, because we do not really judge our fellow humans as thinking beings based on how they answer our questions—we generally accept any human being on sight and without question as a thinking being, just as we distinguish a man from a woman on sight. A conversation may allow us to judge the quality or depth of another’s thought, but not whether he is a thinking being at all; his membership in the species Homo sapiens settles that question—or rather, prevents it from even arising. If such a person’s words were incoherent, we might judge him to be stupid, injured, drugged, or drunk. If his responses seemed like nothing more than reshufflings and echoes of the words we had addressed to him, or if they seemed to parry or evade our questions rather than address them, we might conclude that he was not acting in good faith, or that he was gravely brain-damaged and thus accidentally deprived of his birthright ability to think.

Read the rest at http://www.thenewatlantis.com/archive/11/halpern.htm

Lebenz · #2 02-16-2006, 11:43 AM

Wonder if AC Clarke read Turing and then an early incarnation of the Merck Manual?

Kuan · #3 02-16-2006, 11:48 AM

It's amazing that Turing's model of mind persisted for such a long time. It's appeal was in its elegance. Hey, cut out all that Cartesian dualist nonsense and simplify things right? Yeah..

I wish jjl were here. He'd have something of interest to say, so would djurgba.

It would have been nice to have known the man. The closest I've ever come was attending a lecture by Alonzo Church in the late 80's at UCLA. Imagine, at that time he'd been talking about the same thing for 30-40 years. It never gets old.

kerry · #4 02-16-2006, 02:45 PM

Quote:

Originally Posted by Kuan

It's amazing that Turing's model of mind persisted for such a long time. It's appeal was in its elegance. Hey, cut out all that Cartesian dualist nonsense and simplify things right? Yeah..

I wish jjl were here. He'd have something of interest to say, so would djurgba.

It would have been nice to have known the man. The closest I've ever come was attending a lecture by Alonzo Church in the late 80's at UCLA. Imagine, at that time he'd been talking about the same thing for 30-40 years. It never gets old.

That might have been possible is society had been a little less homophobic.

"In 1952, Turing was convicted of acts of gross indecency after admitting to a sexual relationship with a man in Manchester. He was placed on probation and required to undergo hormone therapy. When Alan Turing died in 1954, an inquest found that he had committed suicide by eating an apple laced with cyanide."

Botnst · #5 02-16-2006, 03:44 PM

Kerry, sexuality is an important issue for you, isn't it?

kerry · #6 02-16-2006, 03:48 PM

Apparently not just me since he was required to undergo hormone therapy.

kerry · #7 02-16-2006, 04:07 PM

Is this a good test question? "How have the hormones effected your sexual desire and how would you compare your inclinations towards homosexual and heterosexual relations before and after?"

A264172 · #8 02-16-2006, 04:19 PM

Quote:

Originally Posted by kerry edwards

"How have the hormones effected your sexual desire and how would you compare your inclinations towards homosexual and heterosexual relations before and after?"

TTuring -(eats apple)

A264172 · #9 02-16-2006, 06:25 PM

Quote:

Originally Posted by Botnst

Turing the mind....

in·tel·li·gence- n.
1)...a. The capacity to acquire and apply knowledge.
......b. The faculty of thought and reason.

I suspect the Turing test is way off track.
What do these intelligent machines do when they aren't being intelligent?
Their intelligence can only be measured astride their purpose.
If they aren't usefull to themselves they will continue to be idiots.

djugurba · #10 02-17-2006, 11:56 AM

I've been taking a refresher course in Developmental Psychology this spring, and it is very interesting how some of the content is very different than when I had taken it previously. Some of the most interesting material we've covered thus far pertains to neuron formation.

Basically (very basically), from stem cells in the blastocyst, the embryo begins to form. Each new cell (and we're talking many many many per minute) gets a specific task, and wanders off to find the other like cells with which to accomplish this task. Neurons have virtual taxi-cab cells that grab them and travel with them to the appropriate part of the forming brain. If they end up in the wrong place, bad things can happen. And, various teratogens introduced at the wrong time can exacerbate these problems. Fetal Alcohol Syndrome is a good example. But, it's pretty amazing to watch (via mega-magnification) newly divided cells feel some sort of fire under their butts to get to somewhere else, along pathways that are still under construction.

As a philosopher, I am always very interested in the ways used to speak of such happenings. More often than not, the language indicates an underlying intentionality on the part of the cells. And so I've been thinking what is involved in generating intentionality.

I can imagine a robot, that, like our neurons, when switched on, is consumed with undertaking a task. Roomba the self-propelled vacuum is one such creature. It is able to accomplish its coded task, with a circuit board in place of a double-helix. We put only the information it needs to accomplish this task inside the roomba, and so it's not very difficult for us to conclude that Roomba does not meet the tests required to be an intentional actor. I'm not sure if neurons are the same, because they've got the entire code, even if they only respond to a little bit of it.

Maybe we could add more coding, perhaps a Bluetooth adapter, and/or a GPS receiver. With this, multiple Roomba(e?) could do what we'd call working together, never overlapping our carpet but rather cleaning their own sections of carpet. Are these coordinated Roombae intentional?

I'm not sure how Turing would handle this. Anything within their scope, we could ask of these Bluetooth Roombae, and they would dutifully perform. We naturally think that there is a difference between drone response to stimulus and actual thought. But I sometimes wonder if a complex enough system (such as, perhaps, our brains) would allow for such subtlety of stimulus response that it might look otherwise. Maybe Turing had it backwards?

Kuan · #11 02-17-2006, 12:52 PM

I think it's easy to confuse the intentionality in the language of description with true intentionality in the system if you're stuck in Tarski correspondence theory model. I, for one, think that language and system are one and the same, and that the object language we use to describe systems are one and the same, ie., the system itself.

Suppose we code something "volitional" into Roomba which makes it turn left at the wall. To the observer with only rudimentary knowledge of the modern world, Roomba wants to turn left when it hits the wall and so we say Roomba seems to want to turn left when it hits the wall. After observing Roomba do it a hundred times it gets predictable. We formulate the general rule "Roomba turns left everytime it hits the wall." The programmer, however, does something else. He could have done something like this:

Rule 1) turn wheels forward until unable to move foward anymore
Rule 2) if unable to move both wheels forward, move wheel 1 until both wheels are able to move forward.

Combined, rules 1 and 2 give us the illusion that Roomba is an intentional system, or even a purposeful system. In otherwords, capable of heuristic decision making.

So what's my point? I ramble a lot I know but I have more than one. My point is that although Roomba is coded as a non intentional being, we can view it as one that is. If it works to explain Roomba's behavior then it should be fine. Roomba is for all intents and purposes an intentional being.

Second point is that we're all victims of our level of understanding. It's our tendency to explain things in terms of cause and effect, in terms of free will and volition, and in the context of some Darwinian struggle for survival. We cannot possibly let Roomba turn left because that's just the way the wheels turn. For some of us it seems impossible that hardware and software are one and the same. We need to lose that outmoded system of thinking.

The big objection my assertion it's all one big thing is that you can change the language and Roomba will do something else. I maintain that yes, you can change the language, but Roomba will not turn left if you tell it to turn right, and we can't use intentional language of any other sort to describe Roomba, so whether you like it or not, the language is part of the whole package. If this sounds to you like a version of Ryle's category mistake argument, then you get the argument.

But Kuan, how can you say that Roomba is an intentional being while at the same time knowing that the code is but a series of statements without any kind of intentionality? You cannot deny that the code exists.

No, I can't. But I also believe that we speak meaningfully about Roomba. We're often realists when we walk around and talk with people but we can also talk and speculate about things which don't exist, and we can do it with meaning. We read novels, we write plays, and we can talk about Roomba as an intentional being without talking in the language of its code.

Botnst · #12 02-17-2006, 01:58 PM

I like that the author of the first article stresses the degree to which appearance affects assessment of intelligence. I think Turing acknowledged that by having the test subject hidden from the tester.

It occured to me that it would be interesting to have a large sample of people who are all told that they are testers then asked to evaluate various entities for the discrimination of natural and artificial intelligence and that they may only communicate via keyboard. Have the large population be various ages, sexes, intelligence levels, language abilities including random native languages.

Then cut them lose by random connections for different lengths of time. Then score the various communicants.

Could they determine something consistently recognizable as 'intelligence' among all of the samples?

Bot

kerry · #13 02-17-2006, 02:36 PM

It's my understanding that hypnotized people, carrying out hypnotically suggested tasks, don't feel compelled to carry out the task, but create their own intentions. Are those intentions any different or less real than everyday intentions? I believe this reinforces Kuan's point.
I also recall reading a study, or a report of a study (perhaps in one of Daniel Dennett's books) in which the initiation of a physical movement of the body occurs microseconds before the mind develops its 'intention' to do so.

A264172 · #14 02-17-2006, 04:41 PM

Quote:

Originally Posted by Botnst

...Could they determine something consistently recognizable as 'intelligence' among all of the samples?...

Maybe they should rank the indications for each answer (I) (not I) and use it as the basis of the program. Then find as many sub-catagories of (I) and (not I) as possible to lend it depth.

Quote:

Originally Posted by djugurba

... But, it's pretty amazing to watch (via mega-magnification) newly divided cells feel some sort of fire under their butts to get to somewhere else, along pathways that are still under construction.
...

Quote:

Originally Posted by Kuan

...The big objection my assertion it's all one big thing is that you can change the language and Roomba will do something else. I maintain that yes, you can change the language, but Roomba will not turn left if you tell it to turn right, and we can't use intentional language of any other sort to describe Roomba, so whether you like it or not, the language is part of the whole package. ...

Quote:

Originally Posted by kerry edwards

It's my understanding that hypnotized people, carrying out hypnotically suggested tasks, don't feel compelled to carry out the task, but create their own intentions...

I find this fascinating. That we obey a thing in our head, because we think it knows what to do.
What tells it what to do?

kerry · #15 02-17-2006, 06:37 PM

I'm thinking that this topic is really the argument for intelligent design in miniature.
We can give a scientific causal explanation for the operations of the brain. There is no end to these explanations.
We stop the empirical causal explanations thru the linguistic convention of the 'self'. "I" am a free responsible agent, determining my actions. There's no empirical evidence of the self, but we attribute agency to the self and stop any further causal explantions by invoking it.
We do precisely the same thing with God and the universe except in that instance there is no internal experience of God giving credibility to the existence of the mysterious cosmic agent. As a result, it is much easier to get rid of the linguistic convention of 'God' than it is to get rid of the linguistic convention of t he self. But despite our internal convictions of responsible agency(especially on the part of those libertarians), there's no scientific evidence we actually exist.