# What is so great about the Turing test? **URL:** \ **Category:** Factual Questions\ **Created:** [December 28, 2008, 9:36pm UTC](https://boards.straightdope.com/t/what-is-so-great-about-the-turing-test/478996 "2008-12-28T21:36:45Z")\ **Posts on this page:** 13\ **Page:** 5
**Author:** ![Indistinguishable](https://avatars.discourse-cdn.com/v4/letter/i/90ced4/32.png) [@Indistinguishable](https://boards.straightdope.com/u/Indistinguishable)\ **Post date:** [January 2, 2009, 7:11pm UTC](https://boards.straightdope.com/t/what-is-so-great-about-the-turing-test/478996/81 "2009-01-02T19:11:01Z")
:smack: Whoops, I’m embarrassed. (I did think “overwhelming misguided agreement” was a bit of an odd thing to say…) ---
**Author:** ![chorpler](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/chorpler/32/2888_2.png) [@chorpler](https://boards.straightdope.com/u/chorpler)\ **Post date:** [January 2, 2009, 8:58pm UTC](https://boards.straightdope.com/t/what-is-so-great-about-the-turing-test/478996/82 "2009-01-02T20:58:44Z")
> [@flaviangens](#): > > If the brain is the material ground of subjective awareness as claimed by materialist philosophy, how does it differ in essence from a digital computer? Apart from ‘I don’t know’, I think the fact it is analogue in nature could be the key. Digital circuits are really analogue but have been constrained by design to respond to threshold voltages, currents and charges. It is this constraint that reduces the computer one or more logical levels from physical nature to the embodiment of an idealisation, and that is why the work it does is essentially (and logically) the same as a pencil and paper dry run of the algorithm it executes, and that may be why the brain can be subjectively aware while digital computers cannot. I’m not sure about this analogue-vs-digital business. Neurons are somewhat digital in nature – in response to whatever the conditions are, a given neuron is either fires or doesn’t fire. There’s no in-between state that I’m aware of … is there? ---
**Author:** ![Indistinguishable](https://avatars.discourse-cdn.com/v4/letter/i/90ced4/32.png) [@Indistinguishable](https://boards.straightdope.com/u/Indistinguishable)\ **Post date:** [January 2, 2009, 9:27pm UTC](https://boards.straightdope.com/t/what-is-so-great-about-the-turing-test/478996/83 "2009-01-02T21:27:30Z")
I’ve basically stayed out of the philosophy and stuck to the math, but all this red herring analog vs. digital discussion makes me want to say: Meh. As far as I’m concerned, quoting Dijkstra, “The question of whether a computer can think is no more interesting than the question of whether a submarine can swim.” When we get to the point where computers act in a manner externally indistinguishable from those beings which are archetypes of “intelligence”, then linguistic norms will settle one way or another as to whether the term gets applied to computers as well (in most contexts or in some contexts or whatever). But it’s not as though it’s some concrete external truth out there for us to discover. It’s just a question of how people employ their language. ---
**Author:** ![flaviangens](https://avatars.discourse-cdn.com/v4/letter/f/5fc32e/32.png) [@flaviangens](https://boards.straightdope.com/u/flaviangens)\ **Post date:** [January 2, 2009, 10:53pm UTC](https://boards.straightdope.com/t/what-is-so-great-about-the-turing-test/478996/84 "2009-01-02T22:53:08Z")
> [@](#): > > Meh. As far as I’m concerned, quoting Dijkstra, “The question of whether a computer can think is no more interesting than the question of whether a submarine can swim.” > > When we get to the point where computers act in a manner externally indistinguishable from those beings which are archetypes of “intelligence”, then linguistic norms will settle one way or another as to whether the term gets applied to computers as well (in most contexts or in some contexts or whatever). But it’s not as though it’s some concrete external truth out there for us to discover. It’s just a question of how people employ their language. Indistinguishable, Dijkstra was showing no imagination. If a program passes the Turing Test we have to decide if we have moral obligations towards it. You may dismiss this also but many will not. If it is decided in the affirmative then a situation could arise that a human is legally punished for abusing a specimen. If it is actually self-aware then it could be just. If is is not then a wrong will have been done. ---
**Author:** ![Alex\_Dubinsky](https://avatars.discourse-cdn.com/v4/letter/a/dbc845/32.png) [@Alex\_Dubinsky](https://boards.straightdope.com/u/Alex_Dubinsky)\ **Post date:** [January 3, 2009, 4:10am UTC](https://boards.straightdope.com/t/what-is-so-great-about-the-turing-test/478996/85 "2009-01-03T04:10:43Z")
> [@Indistinguishable](#): > > :smack: Whoops Hahaha, man. There really is truth to what you were saying. > [@](#): > > What am I totally wrong about? I agree with you that programs of length K can solve the halting problem for all programs of length L, even when L is infinite, as long as K \>= 2^L (and thus K is larger than L). Actually, there’s no reason to bring Cantor’s sets into this. 2^x (where x is infinite) is not the only situation that is \> x. In fact, (and this is where I expect you _will_ disagree), the correct way to formulate infinite numbers is that x+1 is also \> x, as well as 2\*x or x^2. The proofs that claim things like x^2 = x (such as quantity of integers vs quantity of integer tuples) are all based on the confusion between a true “infinite number” and the concept of a “boundless” one. It used to be that the difference between the two was understood, but some prominent early mathematicians overstepped themselves in excitement to claim the two the same. (Back then, a common argument was that infinties do not exist at all, only boundlessnesses. The mathematicians wanted to proclaim infinities exist, which they do. But they went too far and called everything infinite.) Unfortunately, they did not realize the many subtle discrepancies that equating the infinite and the boundless causes. The proofs that x^2 = x are not invalid, but they deal with the case when x is “the boundless concept,” not “an infinite number.” The quantity of integers is boundless, it is not infinite. Boundlessnesses can be multiplied by two and you’ll of course still get something boundless. You can square it, you can add 1, etc, and you’ll still get back to “boundless.” It is not a number. It is a concept. But any actual number, any real infinity, behaves ordinarily. An inf^2, not just a 2^inf, algorithm will also solve the halting problem for inf inputs. ---
**Author:** ![RaftPeople](https://avatars.discourse-cdn.com/v4/letter/r/6f9a4e/32.png) [@RaftPeople](https://boards.straightdope.com/u/RaftPeople)\ **Post date:** [January 3, 2009, 5:17am UTC](https://boards.straightdope.com/t/what-is-so-great-about-the-turing-test/478996/86 "2009-01-03T05:17:23Z")
> [@flaviangens](#): > > Reporting on its ‘thought processes’ would be part of the behavior that lets it pass the Test, but that doesn’t strengthen its claim to have interior, subjective consciousness. In fact the computer is using some data in its memory, along with portions of reference data and the algorithm, to formulate words which report, more or less accurately, on the contents of other, selected (by the algorithm) parts of memory. This is purely a mechanical process, and is wholly deterministic, and predictable in principle by dry-running the program. Do you think this is different from the operation of the human brain? Do you think the brain is not deterministic? > [@](#): > > Because, to paraphrase Thomas Nagel, it is like something to be a subject of experience. It matters to you. It is important to you. Is is the one thing that allows you to _be_. Without it you are _not_. All meaning flows from it and is predicated on it. It is the basis of everything. If that is not intrinsically meaningful, what is? The brain is a computer, it takes external and internal input and produces output. It’s a mathematical function. Same as digital computers although they employ different mechanisms to map input to output. Discussions of morality and subjective experience are interesting at one level of abstraction, but don’t change the underlying fact that it’s just a machine mapping input to output. > [@](#): > > I didn’t mention intelligence, though to pass the Test the program would have to simulate intelligence. No-one knows if the subconscious mind is subjectively aware or not, because its experiences, in as much as they are separate from conscious experience (a very simplistic, and misleading, notion in my opinion), are not directly accessable to the conscious mind. There is more and more evidence that conscious awareness is the “explanation” of what the brain just determined for the benefit of the inwardly pointed observer. The real action appears to be happening before we are aware of it. ---
**Author:** ![Indistinguishable](https://avatars.discourse-cdn.com/v4/letter/i/90ced4/32.png) [@Indistinguishable](https://boards.straightdope.com/u/Indistinguishable)\ **Post date:** [January 3, 2009, 5:47am UTC](https://boards.straightdope.com/t/what-is-so-great-about-the-turing-test/478996/87 "2009-01-03T05:47:43Z")
> [@Alex\_Dubinsky](#): > > Hahaha, man. There really is truth to what you were saying. Phew… _wipes sweat from brow_ > [@](#): > > 2^x (where x is infinite) is not the only situation that is \> x. I agree (barring the Generalized Continuum Hypothesis, which I am not inclined to accept for typical contexts), but I chose 2^x as it would be the number of entries in a database of hardwired answers for every program of size x bits. > [@](#): > > In fact, (and this is where I expect you _will_ disagree), the correct way to formulate infinite numbers is that x+1 is also \> x, as well as 2\*x or x^2. I do not necessarily disagree. Which is to say, there are many different systems we might talk about where it is useful to speak of “infinite numbers”. In typical systems of cardinality, we have that some infinite quantities satisfy x+1 = x. But in systems like the hyperreal numbers, all the usual arithmetic properties continue to hold of the infinite numbers, including x+1 \> x and so on. There is no one-size-fits-all notion of “infinite numbers”, or even of “numbers”. There are different notions for different purposes. > [@](#): > > The proofs that claim things like x^2 = x (such as quantity of integers vs quantity of integer tuples) are all based on the confusion between a true “infinite number” and the concept of a “boundless” one. … An inf^2, not just a 2^inf, algorithm will also solve the halting problem for inf inputs. I might not be as prejudiced towards your views as you suspect. But if you would like to help bring me to a sympathetic understanding, I think you will need to put some effort into explaining the careful details of your views more accessibly. ---
**Author:** ![Frylock](https://avatars.discourse-cdn.com/v4/letter/f/ce7236/32.png) [@Frylock](https://boards.straightdope.com/u/Frylock)\ **Post date:** [January 3, 2009, 11:21pm UTC](https://boards.straightdope.com/t/what-is-so-great-about-the-turing-test/478996/88 "2009-01-03T23:21:39Z")
> [@Indistinguishable](#): > > I’ve basically stayed out of the philosophy and stuck to the math, but all this red herring analog vs. digital discussion makes me want to say: > > Meh. As far as I’m concerned, quoting Dijkstra, “The question of whether a computer can think is no more interesting than the question of whether a submarine can swim.” > > When we get to the point where computers act in a manner externally indistinguishable from those beings which are archetypes of “intelligence”, then linguistic norms will settle one way or another as to whether the term gets applied to computers as well (in most contexts or in some contexts or whatever). But it’s not as though it’s some concrete external truth out there for us to discover. It’s just a question of how people employ their language. It may or may not be just a question of how people employ their language. But even if it is, those kinds of questions can be very important–even matters of life or death in certain cases. It’s a question of _how_ linguistic norms will settle this thing one way or the other. Will there be violence and rioting in the streets, or rational discussion and negotiation, or decision by fiat from the powers that be, or what? We can’t just take a faithful attitude toward the working out that linguistic norms are wont to do–because the working out of linguistic norms is very much something _we_ do. It doesn’t happen independently of us. We participate in that working out. And it seems best to me that we take responsibility for that. I know that seemed a bit melodramatic, and I have no adequate way to answer for that. About some things, I’m melodramatic. 😛 Without assuming there’s a metaphysical fact about sentience that settles how the language ought to be used, I can characterize the problem as follows. I myself tend toward the more “liberal” view that if it _seems_ sentient, and _contributes_ in the same way sentience does, then we ought to treat it as sentient if only to be careful that we don’t accidently kill a fully human-like self-consciousness when we delete its program. And I can adduce non-metaphysical arguments for this position taken from the liberal tradition, concerning the value of diverse ways of life coexisting together and so on. It is _useful_ for me to treat my fellow human beings _not_ as things that ought to be used. If a computer program can be like that in relation to me as well, then there you go–I ought to treat it as though it is not a thing to be used at my whim. Then there could be the more “conservative” view that since it’s not really like us at all, it shouldn’t be regarded as one of us, etc etc. (“Conservative” is really not a fair word to use here, sorry Conservatives!) So that’s kind of a “political” argument that avoids metaphysicalizing the issue by asking if the thing is “really” thinking. But even this “political” argument at least (by being offered) acknowledges that the issue _is_ an interesting and important one, one that it’s worth arguing about. If I don’t offer it, then for all I know, “linguistic norms” will “end up” working things out such that sentient programs are actually only quasi-sentient and should be deleted at our whim. That would be bad. Anyway, that said, I’m not convinced there’s not a job for metaphysics to do here, if _only_ in order to explain why even a radically different (physically) sentience is still _like us_, indeed _metaphysically_, in all the right ways. -FrL- ---
**Author:** ![chorpler](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/chorpler/32/2888_2.png) [@chorpler](https://boards.straightdope.com/u/chorpler)\ **Post date:** [January 4, 2009, 12:13am UTC](https://boards.straightdope.com/t/what-is-so-great-about-the-turing-test/478996/89 "2009-01-04T00:13:10Z")
> [@Frylock](#): > > I myself tend toward the more “liberal” view that if it _seems_ sentient, and _contributes_ in the same way sentience does, then we ought to treat it as sentient if only to be careful that we don’t accidently kill a fully human-like self-consciousness when we delete its program. And I can adduce non-metaphysical arguments for this position taken from the liberal tradition, concerning the value of diverse ways of life coexisting together and so on. It is _useful_ for me to treat my fellow human beings _not_ as things that ought to be used. If a computer program can be like that in relation to me as well, then there you go–I ought to treat it as though it is not a thing to be used at my whim. I completely agree with this point. To point out a recent fictional example among many: in Robert J. Sawyer’s book **Mindscan** , a variant of Searle’s argument is used in court to claim that humans whose minds have been transferred into “mindscan” robots are merely zombies who act like conscious humans, and the mindscans are stripped of their rights to control their property and whatnot. Sure seems wrong to me – if an entity acts like it is conscious, we should treat it as such, I say, unless it can somehow be shown that it really isn’t conscious and experiences no qualia or emotions or whatever. ---
**Author:** ![Indistinguishable](https://avatars.discourse-cdn.com/v4/letter/i/90ced4/32.png) [@Indistinguishable](https://boards.straightdope.com/u/Indistinguishable)\ **Post date:** [January 4, 2009, 12:23am UTC](https://boards.straightdope.com/t/what-is-so-great-about-the-turing-test/478996/90 "2009-01-04T00:23:00Z")
**Frylock** : I guess I can’t disagree with any of that. I mean, we _could_ end up with our moral practices not tightly correlated to our linguistic practices in this area, in the sense that we might end up calling computers “intelligent” but still not feel any qualms about damaging them, or refrain from calling computers “intelligent” but still have taboos on damaging them. But I suppose it’s likely that the two will end generally be correlated to some significant degree. And I agree that discussions like these are part of how linguistic norms get settled. _But_, I guess, the point to me is that, if the only substantive reason we have for being interested in the question of whether to call computers “intelligent” is because of its indirect coupling to our interest in the values of various moral stances regarding computers, then we are liable to confuse our discussion by focusing too much on red herrings. If what we’re really interested in is mores, then let’s discuss mores directly, and not worry about such illusory debates as whether an analog carrier is requisite for proper application of the adjective “intelligent”; that is, let’s not worry about them except insofar as we might have some independent reason to suspect or want to say that whether an entity is analog or digital matters has any moral relevance. ---
**Author:** ![Voyager](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/voyager/32/133_2.png) [@Voyager](https://boards.straightdope.com/u/Voyager)\ **Post date:** [January 4, 2009, 1:02am UTC](https://boards.straightdope.com/t/what-is-so-great-about-the-turing-test/478996/91 "2009-01-04T01:02:01Z")
> [@flaviangens](#): > > Reporting on its ‘thought processes’ would be part of the behavior that lets it pass the Test, but that doesn’t strengthen its claim to have interior, subjective consciousness. In fact the computer is using some data in its memory, along with portions of reference data and the algorithm, to formulate words which report, more or less accurately, on the contents of other, selected (by the algorithm) parts of memory. This is purely a mechanical process, and is wholly deterministic, and predictable in principle by dry-running the program. What if, to simulate choice, it makes some decisions based on a RNG seeded by time of day, exact temperature, etc. You could then reproduce its behavior in a very certain circumstance, but not for all. The question is, how can we tell if someone else has an interior consciousness? Only by applying the test, which we do all the time without thinking about it. I still don’t see how you can tell that a computer does not have consciousness, besides being a priori convinced it does not. Do you think our minds are anything other than a type of process running on our brains? If you do, what is not physical about them? > [@](#): > > Why does it have to be merely ‘a fixed program processing inputs into outputs?’ I said it is an _algorithm_. That means it can do anything that is Turing computational, including modifying the program, and take full advantage of any available memory storage technologies. This includes all possible heuristics. I agree, we could in principle simulate the brain with paper and pencil. At any given time one would be either making a mark on a sheet, erasing a mark, moving a sheet, drinking coffee, sleeping, lying on the beach on holiday etc. Are you proposing that the sum of these activities, crucially with the inclusion of the sheet marking/erasing activity is summoning a spirit? A self-aware being? Located where and when? When is it having a subjective thought? Does it cease to be when you put down your pencil? Does it come back into subjective existence and know itself anew when you pick up your pencil? Are you in fact God that you can do this divine stunt? The “spirit” of this simulation would run wonderfully slowly - but so would the pencil and paper simulation of the computer. I don’t quite get where the coffee enters here. If we simulate a brain, we’d also have to simulate the inputs to the brain - the feeling of hunger and thirst, the smell of coffee, the sight of the coffee. The mapped neurons would already know how to drink, and whether it wants cream or sugar. And where is our self-awareness located? Do you think it is in the brain, or in the soul? I don’t believe in God, so no, I don’t think God has anything to do with it. My kids didn’t seem to be self aware at birth, but grew into it. It seems to be a function of the increasing complexity of our neuronic connections. Is there any evidence for another cause? > [@](#): > > Because, to paraphrase Thomas Nagel, it is like something to be a subject of experience. It matters to you. It is important to you. Is is the one thing that allows you to _be_. Without it you are _not_. All meaning flows from it and is predicated on it. It is the basis of everything. If that is not intrinsically meaningful, what is? To me, maybe. But life existed on this planet for nearly a billion years without consciousness. I’m not sure we have any more intrinsic meaning than that. > [@](#): > > Self-awareness hinging on glitches? A bit of a stretch. Analog computation is fundamentally different from digital computation because the former is continuous while the latter is discreet; it’s the difference between real numbers and integers. And analogue is the way the world is – digital, in the computing sense, is artificial. That is where my point about computers being the embodiment of an idealisation comes from. Anyway, I’m not trying to make a strong case for the analogue/digital distinction being the crucial difference between the embodied brain and the digital computer – it’s just a suggestion. A friend of mine, a famous analog expert, used to laugh at us digital guys and say that when you get fast enough everything is analog. Given Planck time everything is digital at the quantum level. So we may be living in a digital universe after all, like Ed Fredkin said. In any case, connections in our brain have multiple but not infinite levels, so even they aren’t totally analog. The bottom line is I fail to see the real basic difference between a big enough computer, with enough parallelism, and the brain. > [@](#): > > You don’t thing dogs have subjective awareness? They’re zombies, dead inside? That’s rather chauvanistic. I’ve got two dogs, one of whom is a genius with the ability to plan ahead to trick the other out of a bone. But as far as I can tell, neither is aware. Any evidence a dog is. A chip is another story - I’m of two minds about whether they are aware, even mistily aware. > [@](#): > > I didn’t mention intelligence, though to pass the Test the program would have to simulate intelligence. No-one knows if the subconscious mind is subjectively aware or not, because its experiences, in as much as they are separate from conscious experience (a very simplistic, and misleading, notion in my opinion), are not directly accessable to the conscious mind. When you dream, can you direct the dream? I can’t, which I find as evidence that our subconscious is not self-aware. It gives me answers, but never actually communicates, which is more evidence. It all boils down to where you stand on the mind-body problem. If there is a soul, then AIs can never be truly intelligent. If not, I see no reason why they can’t be. ---
**Author:** ![Indistinguishable](https://avatars.discourse-cdn.com/v4/letter/i/90ced4/32.png) [@Indistinguishable](https://boards.straightdope.com/u/Indistinguishable)\ **Post date:** [January 4, 2009, 8:07am UTC](https://boards.straightdope.com/t/what-is-so-great-about-the-turing-test/478996/92 "2009-01-04T08:07:55Z")
> [@Indistinguishable](#): > > And I agree that discussions like these are part of how linguistic norms get settled. _But_, I guess, the point to me is that, if the only substantive reason we have for being interested in the question of whether to call computers “intelligent” is because of its indirect coupling to our interest in the values of various moral stances regarding computers, then we are liable to confuse our discussion by focusing too much on red herrings. If what we’re really interested in is mores, then let’s discuss mores directly, and not worry about such illusory debates as whether an analog carrier is requisite for proper application of the adjective “intelligent”; that is, let’s not worry about them except insofar as we might have some independent reason to suspect or want to say that whether an entity is analog or digital matters has any moral relevance. (Just to head it off, to say “Well, whether an entity is analog or digital might well have moral relevance because _of course_ whether an entity is intelligent or not has moral relevance, while, also, it is plausibly reasonable to limit our application of the term ‘intelligence’ to analog phenomenon” seems to have a tail wagging a dog somewhere. It equivocates on the fact that the “of course” half is a statement about one sense/use of the word “intelligent”, while the “also” half is a note about a narrowing we could choose to undertake, for reasons quite possibly detached from that first sense/use, in carving out another sense/use of the word “intelligent”. That is to say, while there are moods in which we might say the first half, and there are moods in which we might say the second half, we don’t necessarily really mean the same thing by “intelligent” in both of these. Yet, by framing the discussion this way, we risk losing sight of such possible distinctions and context-dependencies. That is the sort of danger I think we run into by spending too much time on questions like “What is the nature of Intelligence?”, to the detriment of many of the applications for which we might have thought these questions to actually be useful.) [It’s like spending a lot of time arguing over whether chess is a “sport”. Well, there are times when it’s reasonable to speak that way, and times when it’s not, and situations where no line has been drawn, but you can draw one if you like, but maybe it’s questionable as to what that really gains you?, and … . Trying to sit down and figure out the essence of “sport”, once and for all, is fruitless. For most applications for which it may be thought that knowing whether chess is a sport might be useful, it is far more productive to forget about drawing unjustifiably sharp boundaries around the word “sport” (whether narrow or broad) and simply tackle the underlying question directly.] ---
**Author:** ![Alex\_Dubinsky](https://avatars.discourse-cdn.com/v4/letter/a/dbc845/32.png) [@Alex\_Dubinsky](https://boards.straightdope.com/u/Alex_Dubinsky)\ **Post date:** [January 4, 2009, 8:58pm UTC](https://boards.straightdope.com/t/what-is-so-great-about-the-turing-test/478996/93 "2009-01-04T20:58:06Z")
Descartes believed in phantasms. He thought the mind was a machine, just balls bouncing around, but on top of that existed something else that did the feeling, the observing. The phantasms did what soulless particles could not possibly do. His idea had a fatal flaw. The phantasms weren’t meant to, couldn’t, affect the balls. They were passive observers, immaterial spirits. If the phantasms (or the observesence, as I called it, or the soul) was the only thing capable of feeling, what did the talking? The balls and gears talked about feeling, but the phantasms were the ones actually doing it? What a disconnect! A machine that goes through the motions, actually feels as well. It must be so. (Although you could argue how many motions it’d have to go through, and that some trivial ones, like throwing out sentences by way of a lookup table, wouldn’t suffice.) [Previous page](https://boards.straightdope.com/t/what-is-so-great-about-the-turing-test/478996.md?page=4)