The Turing P-Zombie
One well-known test for Artificial Intelligence (AI) is the Turing Test in which a test computer qualifies as true AI if it manages to fool a human interlocutor into believing that s/he is having a conversation with another human. No mention of such an AI being conscious is made.
A p-zombie is a being that's physically indistinguishable from a human but lacks consciousness.
It seems almost impossible to not believe that a true AI is just a p-zombie in that both succeed in making an observer think they're conscious.
The following equality based on the Turing test holds:
Conscious being = True AI = P-Zombie
If so, we're forced to infer either that true AI and p-zombies are conscious or that there is no such thing as consciousness.
A p-zombie is a being that's physically indistinguishable from a human but lacks consciousness.
It seems almost impossible to not believe that a true AI is just a p-zombie in that both succeed in making an observer think they're conscious.
The following equality based on the Turing test holds:
Conscious being = True AI = P-Zombie
If so, we're forced to infer either that true AI and p-zombies are conscious or that there is no such thing as consciousness.
Comments (158)
If Bob shoots Joe and it in every way appears motivated by jealousy, does there still remain a possibility that it was not? If you concede there is such a possibility, then you are conceding that behavior is not a perfect reflection of intent, and more importantly, that intent is unobservable.
The point being that behavior does not tell us exactly what the internal state is, which means it's possible that one have a behavior and not have an internal state and it's possible that one have an internal state and have no behavior.
:snicker:
I think it's important to point out that those are two completely different things.
"All" a computer needs to do to pass a Turing Test is say the right words in response to words a judge says. In the way it's typically imagined (there's debate on what a TT really is; let's just ignore that), the thing behind the terminal might be a human or might be a computer. Either way the only access the judge has to he/she/it per the rules is texting via a terminal. So a judge might ask something like, what's a good example of an oblong yellow shaped fruit? And if the response is "A banana", that's something a human could have said. Call that "level 1".
But here's the problem. If we take a "level 1" program and just shove it into a robot, what do you suppose we'd get? It'd be silly to presume you'd get anything other than this... a (hopefully) non-moving robot, incapable of doing anything useful, except possibly over a single channel of communication where it can receive and send sets of symbols equivalent to native language text (e.g. English). If, say, I brought a bowl of fruit, placed it in front of the robot, and typed into this channel, "show me which one is the banana"... then it doesn't matter how well the thing did in the Turing Test, I should expect no action. And that kind of feeble robot is certainly incapable of fooling anyone that it's a human. Before the "oh, that's just a minor detail... suppose we", let's actually suppose we. What do we have to add to this robot to get it to fool a human?
In this example, one thing we might expect the robot to be capable of is, when asked to pick out the banana from the bowl of fruit, that the robot would just reach out and either touch the banana or pick it up. So let's say it does that... then what more is it doing than level 1? Well, it's not just processing string data... now it's observing the environment, associating requests with an action, identifying the proper thing to do when asked to show me which is the banana, and being capable of moving its robot arm towards the banana based on its perception. Now it's not just spitting out words; it's a more semantic thing. It's not just the string "banana" that it has to respond with, but rather, it has to respond with "showing me" (aka, performing some action that counts as "showing me") one of a set of objects (i.e., be capable of identifying where that set is) that is the actual banana (associating the word banana with "the entity" that the word is about). That's a bit more involved than just passing a Turing Test... the two aren't equivalent. You need to do a lot more to build a good p-zombie than just trick people behind a terminal. P-zombies are at least "level 2."
I think that Turing meant that you could have a conversation as opposed to a question/answer session with it. It would be have to access vast amounts of data quickly and come up with the correct sentences, phrases to be able convince anyone that it was a human.
Quoting InPitzotl
I don't remember ever reading anything he wrote or that was written about him that could indicate that he thought AI would be judged by its use in robots. The test was actually based on a game.
Quoting InPitzotl
The observation creates more strings of data for it to process, and make decisions about. Any artificial sense would produce data to be processed. A true AI would have to have a lot of processing power just for that. But for a robot to be able to move you need very little processing power. The two things are not equivalents, AI is not a robot and robots do not have to be AI
Quoting InPitzotl
If I just want to talk to the little black box on my desk because I am lonely, the test works fine. The test does not say that AI has to convince someone by being there in PERSON and convincing them that it is human. That would involve more that just AI, things like appearance, smell, body movement and lots of other human quirks.
I think you're missing the point. Yes, the TT involves having a conversation; but the conversation is limited only to a text terminal... that is, you're exchanging symbols that comprise the language. But the TT involves being indistinguishable from a human to a (qualified) judge. And if your computer program cannot answer questions like this, then it can't pass the TT. Over a terminal, though, all you can do is exchange symbols, by design.
Quoting Sir2u
Mmm.... it's a little more complex than this. Fall back to the TT's inspiration... the imitation game. Your goal is to fool a qualified judge. So sure, if it takes you 10 minutes to figure out that a banana is a good response to an oblong yellow fruit, that's suspicious. If it takes you 10 seconds? Not so much. But if it takes you 5 seconds to tell me what sqrt(pi^(e/phi)) is to 80 decimal places, that, too, is suspicious. You're not necessarily going for speed here... you're going for faking a human. Speed where it's important, delay where it's important.
Quoting Sir2u
I'm not writing a paper discussing Turing; I'm responding to the OP in a thread on a forum. In that post, there was one paragraph talking about an AI passing a TT. The next paragraph, we're talking about p-zombies. All I'm doing is pointing out that these are completely different problem spaces; that passing the TT is woefully inadequate for making you a good p-zombie.
Quoting Sir2u
Technically, yes, but that's a vast oversimplification. It's analogous to describing the art of programming as pushing buttons (keys on a keyboard) in the correct sequence. Yeah, programming is pushing buttons in the right sequence, technically... but the entire problem is about how you push the buttons in what sequence to achieve what goal.
Think of this as skillsets. Being able to talk goes a long way to being a good p-zombie, but it's only one skillset; and mastering just that skill isn't going to fool anyone into thinking "you're" conscious. That "more strings of data" and "decisions about" you're talking about here is another skillset; say, it's a vision analog and you equipped your robot with a camera. That's all well and good, but that skillset is literally about discerning objects (and like "useful things") from the images. English is a one dimensional thing with properties like grammar; images are two dimensional things which convey information about three dimensional objects, which has no grammar... but rather, there are rules to both vision per se and to object behaviors. There's also acting; and reducing that to just moving a part is another oversimplification. Touching a banana isn't a function of "moving a part", but "moving a part appropriately such that it attains the goal of touching the banana"... that involves the vision again, and planning, and the motion, and adjustment, and has to tie in to object recognition and world modeling as well as relating that English text from that first skillset to appropriate actions to respond with. That's another skillset, and it's a distinct one. Furthermore, once your p-zombie starts interacting, it's not truly "just a computer" with "inputs" and "outputs" any more... it's really something more like "two computers" dancing... one with sensors and servos and silicon, but the other, the dance partner, is made of objects, object properties, and physical mechanics. The interaction required to just act towards attaining a goal, a must for fooling a human into thinking you're a conscious agent, is so mechanically interfused with what you're acting with that the entire system becomes important (consider that your actions have effects on your senses and that modeling that, which requires modeling your dance partner, is a required part of the skillset).
Quoting Sir2u
Sure, but that's required to be a p-zombie.
Quoting Sir2u
...well not quite. The p-zombie isn't trying to fool you into thinking that it's a human; it's just fooling you into thinking it's conscious.
So the Turing test is flawed? Behavior is not a reliable indicator of consciousness? Doesn't that mean p-zombies are possible and doesn't that mean physicalism is false?
You are taking the Turing test too literally. The idea of an AI fooling a human has a much broader scope - it includes all interactions between the AI and humans whether just chatting over a network or actual physical interaction.
I can only reply that I've seen people choke on this point. Also, the term Turing Test is a term of art with a literal meaning, so I'm not sure how taking it literally can be a bad thing. p-zombie is also a term of art with a distinct meaning. Surely it's better to just be clear, especially if people get confused, right?
In my humble opinion...
[quote=Bruce Lee]Its like a finger pointing away to the moon. Dont concentrate on the finger or you will miss all that heavenly glory.[/quote]
Alright, let's turn this into a question then. In your original post, you said this:
Quoting TheMadFool
...after which you offered:
Quoting TheMadFool
...so, that reads like it possibly suggests this:
Conscious being = True AI = Passes Turing Test = Fools human interloculators into believing you are having conversations with another human = P-Zombie
So the question is... was that your intent in this thread?
I'm not after a gotcha or a fight; just demonstrating that there's genuine room for confusion here. I'll take your response as a no, so hopefully that would clear things up about your intent.
Are you saying my post is confusing? Well, I did try to keep my wordcount to a minimum. Perhaps that's where the fault lies.
:ok:
Quoting TheMadFool
Never, ever.
[hide]I zom, therefore I be![/hide]
On a serious note, I'm inclined to agree with your last statement from the OP.
Maybe our reports of our own conscious experiences are those of a P-zombie; we're hard-wired to believe that we are conscious..
Silly though it may seem, the notion does hold with the way we work on a neurological level: we have stored memories that exist as a somewhat static record, and every new milliseconds our brains generate new frames of cognition (perception, learning, inference, action).
If we imagine cutting the lifespan of an individual into a series of discrete frames (quite a lot of them I guess) - if we could freeze time -, is a single frame of a live person "conscious"? (by anyone's measure...).
If we merely juxtapose two adjacent frames and flick back and forth between two states (maybe there is some measurable change, like some neurons sending or receiving a new signal), does that create consciousness? Perhaps on the most minimal order of the stuff that consciousness is made of?
The hard problem is pretty hard indeed... Even panpsychism starts to make sense after too long...
I'm tend to air on the side of self-delusion. That consciousness is something "real" (as in, woo woo magic and other presumptive metaphysical rigamarole) is almost certainly delusional. That it's "something" at all beyond a mere report or belief does seem somewhat plausible, but I would not be surprised if we're just self-delusion all the way down.
Someone in another thread had the opinion that consciousness/mind could be an illusion. I take it that he meant there's a physical basis for the phenomenon of consciousness.
This idea of consciousness being an illusion reminds me of Searle's chinese room argument. Searle contends that though true AI will be able to fool a human in conversation, that in itself doesn't prove that it has the capacity to understand like humans. All the AI is doing, for Searl, is mechanically manipulating symbols and the AI never actually understands the meanings of the symbols and their combinations.
That raises the important question of what understanding is and, more importantly, whether it is something beyond the ability of a computer AI? Speaking from my own experience, understanding/comprehension seems to start off at the very basic level of matching linguistic symbols (words, spoken or written) to their respective referents e.g. "water" is matched to "cool flowing substance that animals and plants need", etc. This is cleary something a computer can do, right? After such a basic cognitive vocabulary is built what happens next is simply the recognition of similarities and differences and so, continuing with my example, a crowd of fans moving in the streets will evoke, by its behavior, the word and thus the concept "flowing" or a fire will evoke the word/concept "not cold" and so on. In other words, there doesn't seem to be anything special about understanding in the sense that it involves something more than symbol manipulation and the ability to discern like/unlike thing.
If we were to agree with Searle on this then the onus of proving understanding/comprehension is not just symbol manipulation falls on Searle's supporters.
:smile:
What does it mean to be physically indistinguishable? Are there other ways of being distinguishable or indistinguishable?
Quoting TheMadFool
I don't know. What is "consciousness"?
Good question but how might I word it to be more explicit than that? Perhaps physical in the sense that the p-zombie has a head, trunk, limbs, internal organs - identical in every sense of bodily parts?
Quoting Harry Hindu
What is consciousness? Perhaps best elucidated as the difference between sleep and awake states.
My computer goes to sleep sometimes and then wakes up when I move the mouse or hit a key on the keyboard.
What if someone is dreaming? Are they conscious?
Indeed, what else could "physical" mean?
Quoting Harry Hindu
Could the computer be conscious?
Quoting Harry Hindu
You overlooked non-REM sleep.
Waking and sleeping states aren't physical states?
Quoting TheMadFool
Well, you did define consciousness as the difference between waking and sleeping states, so it seems to be the case, yes.
I want to clarify what consciousness is.
Also, what are alseep and awake states then, if not physical?
Quoting Harry Hindu
So, a standard issue computer is capable of consciousness? I guess we're not seeing eye to eye on what consciousness means.
Why don't you give it a go? What is consciousness to you?
It's not you and I that aren't seeing eye to eye. You aren't seeing eye to eye with your previous statement.
What makes it impossible for a "standard issue computer" to be capable of consciousness if you defined consciousness as the difference between waking and sleeping states? If it is still impossible even though you defined it as such, then is consciousness something more than just the difference between waking states, or something else entirely that has nothing to do with waking and sleeping states?
Quoting TheMadFool
I got to what consciousness is for me by asking these questions that I'm asking you to myself. I think that if I tell you what I think consciousness is, it would turn into an argument. Let's see where these questions lead us.
Quoting TheMadFool
Then why are you trying to determine if consciousness exists by distinction in body type and function, rather than being awake or asleep? I could build a humanoid robot that goes to sleep and wakes up, like a "standard issue computer". Is it conscious? If P-Zombies look and behave like humans, which includes going to sleep and waking up, then p-zombies are conscious.
When I mentioned sleep and awake states I thought you'd immediately know that the domain of discussion is humans and not anything else.
Quoting Harry Hindu
I'm all eyes and ears. We disagree. I'd like to know why and for that I need to know what you think consciousness is.
I think that's part of the problem - anthropomorphism.
I thought you were talking about p-zombies too, and the point still applies to them:
Quoting Harry Hindu
That's begging the question.
I asked you this:
Quoting Harry Hindu
If there is something more, then what is it - that it has to be a human? Then by definition p-zombies aren't conscious because they aren't humans.
I didn't provide a definition. If I did anything, it's give you just a rough idea of what I think consciousness is. Again, with the hope of coming to some agreement with you, what is your definition of consciousness?
Then you mistook what I was asking for. I wasn't asking for a rough idea, but a specific one as you seemed to know the specifics if you can behave like the arbiter of what is conscious and what isn't. If you've already determined that you must be a human to be conscious, then you've answered your own question.
Your qualifiers were waking/sleeping and being human. P-zombies fit the former but not the latter, therefore p-zombies being conscious is false.
If you're going to restrict the discussion to only humans then you're not going to agree with my definition, but then that would exclude p-zombies from the discussion as well, and your thread is inadequately named.
The original game was to try to distinguish between a man and a women using only written text answers to questions, obviously speaking would have made it too easy.
I think that you missed the point of exactly what he said. Look it up.
Turing did not set any limit to how the test would be carried out with a computer, his vision of a computer that was capable of fooling a human was 50 years from his time. I doubt that he thought that computers would just stay as text terminals, even though they did not exist in his time. His statement was futuristic.
Quoting InPitzotl
You can program response time and delays into a computer quite easily, an AI would have a basic set of rules to follow for most of its operations. Just as it would need rules to follow when picking something up.
Quoting InPitzotl
So an AI would need trillions of lines of code instead of millions, which brings us back to the part of processing power. Sequences using IF/THEN/ELSE would decide which buttons were pressed and when, again trillions of them.
Quoting InPitzotl
Computer learning has come a long way in the last few years. Some of them can and do recognize objects. Some of them can and do pick them up and manipulate them, with great competence. Put the two together and there is your machine.
While accomplishing this has not happened in the 50 years Turing predicted it is getting closer everyday.
But one thing that most people seem to forget about when talking about the theoretical P zombie is that it is not written anywhere that it has to be a physical object, only that it has to be able to convince real people that it is a real person. A hologram would probably be able to do that.
If someone throw a banana at you, would you try to catch it or dodge it. Fear of contact from a physical object would be just as convincing as actually catching it.
Sure. Here's a pdf copy; and here's an html one. (Context for others... these are links to Alan Turing's article "Computer machinery and Intelligence", 1950, which introduced what's now regarded as the Turing Test).
Quoting Sir2u
That's reasonable; since it would involve more things, it likely would involve more code.
Quoting Sir2u
Well, yeah, it has.
Quoting Sir2u
"Forget" is a strong word; that implies not remembering something said. Cite?
Sorry about that, definitely badly worded. I have been busy in classes all day and just skipped a few minute to drop in here at lunch. Let me rewrite that part.
But one thing that most people don't seem to realize when talking about the theoretical P zombie is that it is not written anywhere that it has to be a physical object, only that it has to be able to convince real people that it is a real person. A hologram would probably be able to do that.
Is that the only problem?
Firstly, why are you so coy about your definition of consciousness?
Secondly, I'd like to know what your analysis of the Turing test is vis-a-vis consciousness and p-zombies? The Turing test would have us believe that behavior alone (of the AI) suffices to come to the conclusion that the AI is conscious. Compare and contrast that to the p-zombie in which case, if a p-zombie is possible, behavior alone is insufficient to infer consciousness.
The problem with p-zombies is that they can debate consciousness in just as nuanced a manner as a philosopher like Chalmers or any of us discussing our everyday subjectivity. I find that bordering on incoherent.
So my assumption would be that if an AI can pass a robust Turing Test on consciousness, then it's probably conscious like us. But it has to be robust, and not just clever programming techniques. Humans are easily fooled since we have a tendency to see agency in things.
The characteristic of consciousness is not a good line in bullshit, but giving a fuck.
So anyone not blinded by their own bull, has no difficulty recognising the consciousness of someone with Downs, or a cat or a bird. Philosophers and world leaders are the difficult borderline cases.
But why is it "bordering on incoherent"?
What do you mean?
"Siri, what's the temperature?"
"It's 20 degrees outside. Brrrr, cold."
Think about a p-zombie telling other p-zombies about a dream they had. Now what could the dream teller mean, and what would the listeners understand, given that they have no dream experiences?
Or take a movie like one of the Terminator ones where at some point you see the world from the first person perspective of the terminator. Now how would a p-zombie understand that?
Starting from the fact that we cannot agree on just what is consciousness and have big problems in deciding just what is and what isn't conscious, it's hardly surprising that even a brilliant mind like Turing would be vague on the subject.
After all, a far more simple and theoretical issue like calculation, computability / uncomputability puzzles us still quite a lot.
Yet the the proponents of computers and computer theory have been very willing to declare AI to be actuality even now, whereas many laymen still consider real AI to be that truly conscious robotic chap that indeed has a mind of it's own. Needless to say, a smart program can pass the Turing test many times. But with luck a clever recording would do that also sometimes...
I guess the point was: something is intelligent, if it is called intelligent because it appears to be intelligent.
The judgement is already made by people, not by some arbitrary criteria. :lol:
The written form of communication was chosen to prevent a bias based on seeing a human or not.
One of the best points is made in Gibsons "Neuromancer"
Case: "Are you sentient?"
AI: "Well, if you ask me then I am. But I guess that is some kind of philosophical question. hrhr"
Intelligent?
Because you've restricted the domain of the discussion to humans.
One of the possible answers to "what is the definition of consciousness" is, "I dont know". From there we dont assume that it is necessary to be human to be conscious.
Quoting TheMadFool
What type of behavior is indicative of being conscious? Any human behavior? What about sleeping?
And what did we end up concluding here:
Quoting TheMadFool
If there are no other ways for something to be distinguishable or indistinguishable, then "physically" is a useless term, at least in the context of the distinguishable and indistinguishable.
Maybe the dichotomy between the physical and mental is just as useless and should be considered when defining consciousness.
Is consciousness something you have, something you do, or something you are?
The reasoning in Turing's test is quite similar to yours: that we'd just notice it, because we are conscious. Yet the fact is that appearances can be deceptive.
del
Are you saying that certain behavior, e.g. talking about dreams, thoughts, feelings, etc, is sufficient to infer the presence of consciousness? If so, how do you reconcile your point of view with the Turing test which basically claims that all a computer has to do is mimic a person, converse with us as if it has dreams, feelings and thoughts while not actually experiencing them? I mean the Turing test suggests, in no uncertain terms, that consciousness isn't necessary to be a human.
Into what other areas would you like to extend this discussion into? I'm game.
If Turing thought that a computer AI has only to mimic a human to qualify as conscious then it seems he would also think the p-zombies are conscious.
To say that a computer mimics a person is already defining consciousness as something that can simulated or emulated. Can consciousness be mimicked or is it that wherever some behavior exists consciousness necessarily exists and can't be something that is mimicked?
It seems that Turing thought that behavior of an entity is adequate grounds to believe that entity to be conscious. Chalmers, because he thinks p-zombies are possible, doesn't share Turing's sentiments.
Did he have that in mind? I haven't read his papers well enough to make that specific conclusion. If you have a direct quote, feel free to enlighten me.
Because, again, how do I know I'm not responding to a very clever bot here either, but another human being?
I can make the argument that your responses seem to be made by a conscious human being. But that assumption doesn't mean I think p-zombies are conscious. What I do know is that we don't understand consciousness yet, simple as that.
If you think, as Turing supposedly did, that consciousness can be inferred from the behavior of a computer then it isn't much of stretch to conclude that Turing would've come to the conclusion that p-zombies are impossible.
Turing wasn't concerned with philosophy. And without reading his papers I guess he spoke of intelligence, not consciousness.
If p-Zombies make you headache, may I ask, do you think other people are zombies or not and why? This surely would make for a life-style...
I'm afraid reading his papers isn't something I've done; I wouldn't understand it anyway. Do you suppose Turing wasn't talking about consciousness when he formulated the now famous Turing test?
I find that hard to believe because the test specifically mentions that the AI has to convince a human that it (the AI) is human and being human involves consciousness - in fact consciousness is the defining feature of being human.
You mentioned "intelligence". Going with that, AI would have to mimic human intelligence. See below:
[quote=wikipedia]Human intelligence is the intellectual capability of humans, which is marked by complex cognitive feats and high levels of motivation and self-awareness[/quote]
Self-awareness is the cornerstone of consciousness.
I guess this is more about transitivity. Humans are assumed to be intelligent. Commonly this is assumed to be indicated by intelligent behaviour and communication. Therefor the behaviour of an intelligent machine must be indistinguishable from human behaviour in this respect.
It is not really about pretending to be a human.
What do you mean?Quoting Heiko
Is this a false assumption?
Quoting Heiko
The Turing test specifically states that all the AI has to do is give the impression that it's a human. Pretend?
If two subject are the same and one of them is the same as yet another, then all three are the same.
Quoting TheMadFool
It is anthropocentric.
Quoting TheMadFool
If it can it's behaviour must be intelligent.
I would guess that (due to science-fiction) today's society is much more open to the idea of an intelligent machine than maybe was the case at Turing's time. I guess this is reflected in the whole setting. Today there are software engineers running around trying to tweak natural language processing as they know where the flaws are. I would take it pretty seriously if they once say: "Okay, now we really cannot tell the difference anymore."
So put it clear: We want computer-slaves that can solve our problems without us doing much work.
What are the risks? What is the potential profit worth?
It seems the last thing "philosophers" would consider is asking the machine if it had any problems with doing so. That is significant
Can you give me some examples?
Quoting Heiko
Kindly elaborate.Quoting Heiko
You would and even I would (probably) but how would one know whether a computer is conscious in the sense we are?
With regard to your specific claim that there can be behavior without internal state: certainly, but once you have observed more than a few behaviors that do seem to be dependent on internal state (e.g. learned behavior, or any behavior apparently using memories), then the possibility that none of the observed behaviors were actually state-dependent becomes beyond-astronomically improbable.
Do you know that I am conscious in the same way that you are? (or that any other person is, for that matter.) If so, then apply whatever method you used to come to that conclusion to a computer - and if that method depends on me being human and is not applicable to computers, then you would be begging the question.
Haven't been following discussion but noticed this. If I may jump in, I would say that this is THE core question.
I put it in the following even stronger form: How do I know my neighbor is self-aware and not a p-zombie? I leave the house in the morning and see my neighbor. "Hi nieghbor." "Hi ff. Looks like a nice day." "Yeah sure is." "See you later." "You too!" I cheerfully respond as I drive off with a wave. What a sentient fellow, I think to myself. He surely must possess what Searle would call intentionality. This is my little joke to myself, how little evidence we accept for the sentience of people. The interrogator is always the weak link in real life Turing test experiments. The humans are always TOO willing to accept that the computer's a human.
In truth, I operate by the principle that my neighbor is a fellow human, whatever other differences we may have; and that all fellow humans are self-aware. That's my unspoken axiom.
Computer scientist and blogger Scott Aaronson calls this attitude meat chauvinism and he has a point.
I have no way of knowing if my neighbor is self-aware, let alone some inanimate program. But at the same time I must admit that just because a thing is different from me, does not count as evidence that the thing is not intelligent. If self-awareness can arise in a wet messy environment like the brain; why shouldn't it arise in a box of circuit boards executing clever programs?
Personally I don't think machines can ever be conscious; but still I do admit my human-centric bias. I have no proof that self-awareness can only arise in wetware. Who could begin to know a thing like that? The wisest among us don't know.
And of course this was Turing's point. As a closeted gay man in 1950's England, he argued passionately for the acceptance of those who were different. That's how I read his 1950 paper, not only mathematically but also psychologically.
If we define self-awareness as a purely subjective experience, then by definition it is not accessible to anyone else. There is no hope of having an actual self-awareness detector. Turing offers an alternative standard: Behavioral. If we interact with it and can't tell the difference, then there is no difference.
Some days I do wonder about my neighbor.
@fishfry
The difficulty with employing a method to detect consciousness is that such a method is completely behavior-dependent and that raises the specter of p-zombies, non-conscious beings that can't be distinguished from conscious beings behaviorally.
The Turing test, as a way of identifying AI (conscious), simply states that if there's no difference between a candidate AI and a human in terms of a person who's assessing the AI being fooled into thinking the AI is human then, the AI has passed the test and, to all intents and purposes, is conscious.
P-zombies are treated differently: even if they pass the Turing test adapted to them, they're thought not to deserve the status of conscious beings.
In short, something, AI, that's worlds apart from a flesh-and-bone human, is considered worthy of the label of consciousness while something that's physically identical to us, p-zombies, aren't and in both cases the test for consciousness is the same - behavior based.
Yes.
Quoting TheMadFool
Yes but with a big caveat. You used the word conscious but Turing uses the word intelligent. Intelligence is a behavior and consciousness is a subjective state. So we could use the Turing test to assert that we believe an agent to be intelligent, while making no claim about its consciousness. In fact asking if an agent other than yourself is conscious is essentially meaningless, since consciousness (or self-awareness) is a subjective state.
So that's a semantic difference. The Turing test evaluates intelligence (whether it does that successfully or not is a matter for discussion). But it makes no claims to evaluate consciousness nor do we think any such test is possible, even in theory. Not for an AI and not for my neighbor, who's been acting strangely again.
Quoting TheMadFool
Thought by whom? You and me, I think. But Turing and Aaronson would say, "If it acts intelligent it's intelligent. And nobody can know about consciousness."
On what rational basis do we say our neighbors are conscious but that a general AI, if one ever passed a more advanced and clever version of the Turing test, is "intelligent but not conscious." That sounds like an unjustified value judgment; a prejudice, if you will. Meat chauvinism.
Quoting TheMadFool
Isn't that just an irrational prejudice? They used to say the same about certain ethnic groups. It's the same argument about dolphins. Just because they don't look like us doesn't mean they're not "considered worthy of the label of consciousness." What is your standard for that?
I hope you see the problem here. If we can't tell a human from a p-zombie then what's the difference? I'm not advocating that argument, I disagree with it. I just admit that I can't articulate a rational defense of my position that doesn't come down to "Four legs good, two legs bad," from Orwell's Animal Farm.
I realize that AI = artificial intelligence and not consiousness.
Firstly, even if that's the case, we still have the problem of inferring things about the mind - intelligence, consciousness, etc. - from behavior alone. The inconsistency here is that on one hand (AI) behavior is sufficient to conclude that a given computer has a mind (intelligence-wise) and on the other hand, p-zombies, it isn't (p-zombies have no minds).
Secondly, my doubts notwithstanding, intelligence seems to be strongly correlated with consciousness - the more intelligent something is, the more capacity for consciousness.
In addition, and more importantly, aren't computers more "intelligent" in terms of never making a logical error i.e. Turing had something else in mind regarding artificial intelligence - it isn't about logical thinking which we all know for a fact that even run of the mill computers can beat us at.
What do you think this something else is if not consciousness? Consciousness is the only aspect of the mind that's missing from our most advanced AI, no? The failure of the best AI to pass the Turing test is not because they're not intelligent but because they're not, or are incapable of mimicking, consciousness.
In short, the Turing test, although mentioning only intelligence, is actually about consciousness.
Quoting fishfry
It's not meaningless to inquire if other things have subjective experiences or not.
Quoting fishfry
All I'm saying is a p-zombie is more human than a computer is. Ergo, we should expect there to be more in common between humans and p-zombies than between humans and computers, something contradicted by philosophy (p-zombie) and computer science (Turing test).
If many humans cannot pass the Turing test, is it really a good test?
Quoting TheMadFool
Firstly, just to be clear, and as you say in your original question, p-zombies are imagined as not merely behaviorally indistinguishable from humans, but entirely physically indistinguishable (and therefore beyond scientific investigation - and if they could be studied philosophically, I don't think anyone, not even Chalmers, has explained how.)
Secondly, I don't think the Turing test should be considered as the only or ultimate test for consciousness - it was merely Turing's first shot at such a test (and, given that, it has endured surprisingly well.) For the purpose of this discussion, we can use Turing 's version to stand in for any such test, so long as we don't get hung up on details specific to its particular formulation.
I assume that you think other people are conscious, but on what is your belief grounded? Is it because they behave in a way that appears conscious? Or, perhaps, is there an element of "they are human, like me, and I am conscious?"
If you are going to throw out all behavioral evidence, in the case of AI, on account of p-zombies, then you would be inconsistent if you did not also throw it out in the case of other people. If you make an exception for other people because they are human, then you would be literally begging Chalmers' 'hard question of consciousness'. What does that leave? If you have another basis for believing other people are conscious, what is it and why would that not work for AIs? Suppose we find an enclave of Neanderthals, Homo Erectus, or space aliens - how would you judge if they are p-zombies?
This, I suspect, is what Dennett is alluding to when he says "we are all p-zombies" - he sees no reason to believe that we have these extra-physical attributes that p-zombies would lack.
Returning to your original question, you raise an interesting point: why should p-zombies not be considered conscious? After all, they were conceived explicitly to be indistinguishable from conscious entities. That they are allegedly lacking something that a philosopher says an entity must have, to be conscious, is not much of an argument; the philosopher might simply have an inaccurate concept of what consciousness is and requires.
Putting that aside, there is a third option to be considered: that p-zombies are ultimately an incoherent concept. When we look at how strange, ineffable, unique and evidence-free a concept Chalmers had to come up with in order to defeat physicalism, how selective he had to be in what he chooses to bless with being conceivable, in order to get there, and the highly doubtful leap he makes from conceivability to possibility (I can conceive of the Collatz conjecture being true and it being false, but only one of these is possible), I am simply going to apply Occam's razor to p-zombies, at least until a better argument for their possibility comes along.
What test do you propose? Any ideas?
Quoting A Raybould
Yes, that is the sticking point. I don't see a light at the end of this tunnel.
Quoting A Raybould
What, according to you, is an "accurate" concept of consciousness?
Quoting A Raybould
Why is it incoherent? Quoting A Raybould
How and where is Occam's razor applicable?
I think we're agreed on that.
I jumped into this thread based on only one phrase from one post without reading the rest. I only posted to get out of my system pretty much everything I knew about the subject. I have no idea what the answers are to the question of consciousness and machines; but I do think I have a fair grasp of the questions, at both the technical and philosophical level.
So I said my piece, and if it's not clear, I'm not espousing or even expressing any kinds of opinions about anything. If you disagree with anything I write, that's perfectly ok. I disagree with a lot of it too. I probably won't engage though. I literally said, at a high level, everything I know abut the philosophy machine intelligence in my first post.
Quoting TheMadFool
Yes ok. I'm not agreeing or disagreeing or having an opinion about this, beyond what I've already said. I perfectly well agree with your analysis of the problem.
Quoting TheMadFool
AHA! But ... why do you say that? Until you give a rational reason for WHY you believe that to be the case, I regard it as an entirely bio-centric prejudice on your part. Mean chauvinism again. So on that point, I am pushing back a bit on your ideas. I want to know what is the rational reason we think that intelligence must correlate with consciousness? Other than that it's how our mind works?
Quoting TheMadFool
No, Gödel and Turing decisively delineated the hard limitations of what can be computed. The core of the argument that humans do something machines can't do is that WE can solve noncomputable problems. Sir Roger Penrose believes consciousness is not a computation. I personally believe consciousness is not a computation.
Computers aren't intelligent at all. They're dumb as rocks. In fact you can implement a computer with a bunch of rocks painted white on one side and black on the other. You can sit there flipping rock-bits according to programs, and what you are doing is computing. Computers don't know the first thing about anything. That's my opinion. Others have different opinions. That's ok by me.
Quoting TheMadFool
Nonsense. Turing never used the word. You're adding your own interpretation to what's not actually there. Do you know anything about how chatbots work? People have a tendency to think dumb chatbots are human. That means nothing.
Quoting TheMadFool
It's not meaningless, it's just damned hard to investigate! I heard one thing they do is the "mirror test." If an animal or a fish or whatever can recognize its own reflection, we think it's got some kind of inner life going on.
I don't disrespect or downplay the importance of the question.
I do oppose overly glib and strongly asserted answers. Truth is nobody knows.
Quoting TheMadFool
As I mentioned I haven't read the rest of the thread and wasn't really talking about p-zombies except as a shorthand for Turing machines passing as humans among us in society. Essentially the same idea as the standard meaning of "something that looks and acts exactly like a normal person, but has no inner life at all."
For purpose of anything I'm saying, these two ideas of p-zombies can be taken as the same. I'm not really up on any fine points of difference. A standard p-zombie looks human but has no inner life. A Turing machine operating a perfectly lifelike robot body would in my opinion BE a p-zombie; but I guess you'd say that if it behaves with intelligence, it must be conscious.
Quoting TheMadFool
I'm afraid I don't see the distinction between p-zombies and computers. In my opinion a program running on a computer might possibly implement a true p-zombie -- something that behaves perfectly like a human; but that has no inner life whatsoever.
If all you mean is that the p-zombies are wetware, why do you have such a meat prejudice? Where does it say that being made of meat is better than being made of vegetable matter? It's meat chauvinsim: believing that meat is superior because we are meat. That is not a rational argument.
Quoting TheMadFool
I am not aware of this interpretation, but I don't know much about p-zombies. It seems to me that a lifelike chatbot is exactly what philosophers mean by a p-zombie: a thing that behaves like a human but isn't and that has no inner life. It just operates on a program. Like a computer.
I see p-zombies and computer programs as being very closely related. Perhaps you can educate me as to what I'm missing about p-zombies.
I once half-jokingly suggested that devising a test that we find convincing should be posed as an exercise for the AI. The only reason I said 'half-jokingly' is that it would have a high false-negative rate, as no human has yet completed that task to everyone's satisfaction!
I do not think Turing, or anyone else until much later, anticipated how superficially convincing a chatbot could be, and how effectively a machine could fake the appearance of consciousness by correlating the syntactical aspects of a vast corpus of human communication. By limiting his original test to a restricted domain - gender roles and mores - Turing made his test unnecessarily defeasible by these means, and subsequent variations have extended the scope of questioning. These tests could be further improved by focusing on what a machine understands, rather than what it knows.
While there is a methodological difficulty in coming up with a test that defeats all faking, this is not the same problem as p-zombies allegedly pose, as that takes the form of an unexplained metaphysical prohibition on AI ever being 'truly' conscious (by 'unexplained', I mean that, in Chalmers' argument, we merely have to conceive of p-zombies, without giving any thought to how they might be so.)
Quoting TheMadFool
I don't know, any better than the next person, what consciousness is, and if anyone had come up with a generally-accepted, predictive, falsifiable explanation, we would no longer be interested in the sort of discussion we are having here! For what it's worth, I strongly suspect that, for example, theories linking consciousness to quantum effects in microtubules are inaccurate. In a more general case, I think that any argument that insists physicalism must require a purely reductive explanation of conscious states in terms of brain states, without considering the possibility that the former may be emergent phenomena arising from the latter, are also inaccurate.
Quoting TheMadFool
I am not saying that p-zombies are definitely an incoherent concept, though I suspect they are - that it will turn out that it is impossible to have something that appears to be as conscious as a human without it having internal states analogous to those of humans.
Chalmers defends p-zombies as being "logically conceivable", but that is a vary low bar - it simply means that it is not simply a self-contradictory concept (such as 'male vixen' is, to quote one of Chalmers' examples), and that we don't know of any fact that refutes it - but that is always at risk of being overturned by new evidence, as has happened to many other concepts that were once seen as logically conceivable, such as phlogiston (actually, some form of metaphysical phlogiston theory might still be logically conceivable, but no-one would take seriously an argument based on that.)
Quoting TheMadFool
Chalmers is looking forward to a time when neuroscience has a thorough understanding of how brains work, and is trying to say that no such explanation can be complete - that there is something non-physical or magical going on as well. He cannot say what that is or how it works, or offer any way for us to answer those questions, but he insists that it must be there. It is for exactly these sort of unfalsifiable claims that Occam's razor was invented (even though the concept of falsifiability was not explicitly recognized until centuries later!)
Chalmers' canonical p-zombie argument is a mataphysical one that is not much concerned with computers or programs, even though they are often dragged into discussions of AI, often under the misapprehension that chatbots and such are examples of p-zombies. The argument is subtle and lengthy, but I think this is a good introduction.
That solves the mystery of who or what we consider to be more "intelligent"? I think, despite the who's who of computer science constantly reminding us that computers are not intelligent, people are still under the impression that computers are. I wonder why? Even you, presumably a person in the know about the truth of computer "intelligence", half-thought they were suited to a task humans have difficulty with.
Quoting A Raybould
I see nothing special in understanding. For the most part it involves formal logic, something computers can do much faster and much better. I maybe wrong of course and you'll have to come up with the goods showing me how and where.
Quoting A Raybould
If I maybe so bold as to hazard a "proof": The idea that certain behavior is adequate grounds to infer consciousness is, I believe, an inductive argument. Each person is aware that s/he's conscious and that this state has a set of behavioral patterns associated with it. S/he then observes other people conduct themselves in a similar fashion and then an inference of consciousness is made. Nevertheless, this association between consciousness experienced in first person and some set of behaviors is not that of necessity (no deductive proof of it exists) but is that of probability (an inductive inference made from observation). Ergo, in my humble opinon, p-zombies are conceivable and possible to boot.
Can you summarize the main point please? What is a p-zombie if it's not a TM (or some alternate mechanism) that emulates a human without being self-aware? I confess that I have little inclination to dive into a subtle and lengthy piece by Chalmers. My limitation, I admit. The question I asked is simple enough though. What's a p-zombie if not a TM or super-TM (TM plus some set of oracles) that emulates a human without being self-aware?
ps -- Feeling guilty at my laziness for not clicking on your link, I clicked. And it's NOT an article by Chalmers at all. I skimmed quickly but didn't read it. I'd still like a simple answer to how a p-zombie differs from "a thing that is indistinguishable from a human but lacks self-awareness," such as a TM in a nice suit.
No, it is intended to be what you asked for, an alternative to the Turing test, and the purpose of that test is to figure out if a given computer+program is intelligent.
Quoting TheMadFool
I don't see where you got that from. I am writing about a hypothetical future computer that at least looks like it might be intelligent, just as Turing was when he presented his test.
Quoting TheMadFool
If that is so, then how come the most powerful and advanced language-learning program has a problem with "common-sense physics", such as "If I put cheese in a refrigerator, will it melt?"
Consider Einstein's equation E = mc^2. A great many people know it, but only a tiny fraction of those understand how it arises from what was known of physics at the beginning of the 20th. century. Einstein did not get there merely (or even mostly) by applying formal logic; he did so through a deep understanding of what that physics implied. A computer program, if lacking such insight, could not find its way to that result: for one thing, the problem is vastly too combinatorially complex to solve by exhaustive search, and for another, it would not understand that this formula, out of the huge number it had generated, was a particularly significant one.
Quoting TheMadFool
For one thing, you seem to be making an argument that they are conceivable, but the controversial leap from conceivable to possible is not really argued for here, it is just asserted as as if it followed automatically: "...and possible to boot."
More interestingly, if I am following you here, you do consider it possible that other people are p-zombies. That is very interesting, because Chalmers hangs his argument against physicalism on the assumption that they are not, and I know of no counter-argument that challenges this view (even when Dennett says, apparently somewhat tongue-in-cheek, that "we are all p-zombies", I think his point is that he thinks Chalmers' distinction between p-zombies and us (the non-physical element that they lack) is illusory.)
Having said that, I have three follow-up questions: firstly, if other people could be p-zombies, do you think that you are different, and if so, why? Secondly, if it is possible that other people are p-zombies, why would it matter that it would be possible for a p-zombie to pass the Turing test? Thirdly, if it is possible that other people are p-zombies, why did we evolve a highly-complex, physical state machine called the human brain? After all, if the p-zombie hypothesis is correct, our minds are independent of the physical brain. The most parsimonious hypothesis here seems to be that the p-zombie hypothesis is false, and our minds actually are a result of what our physical brains do.
Indeed, that article is not by Chalmers; is that a problem? Is reading Archimedes' words the only way to understand his principle?
If you want to read Chalmers' own words, he has written a book and a series of papers on the issue. As you did not bother to read my original link, I will not take the time to look up these references; you can find them yourself easily enough if you want to (and they may well be found in that linked article). I will warn you that you will find the papers easier to follow if you start by first reading the reference I gave you.
Quoting fishfry
That is a different question than the one you asked, and I replied to, earlier. The answer to this one is that a TM is always distinguishable from a human, because neither a human, nor just its brain, nor any other part of it, is a TM. A human mind can implement a TM, to a degree, by simulation (thinking through the steps and remembering the state), but this is beside the point here.
If you had actually intended to ask "...indistinguishable from a human when interrogated over a teletype" (or by texting), that would be missing the point that p-zombies are supposed to be physically indistinguishable from humans (see the first paragraph in their Wikipedia entry), even when examined in the most thorough and intrusive way possible. This is a key element in Chalmers' argument against metaphysical physicalism.
As a p-zombie is physically identical to a human (or a human brain, if we agree that no other organ is relevant), then it is made of cells that work in a very non-Turing, non-digital way. Chalmers believes he can show that there is a possible world identical to ours other than it being inhabited by p-zombies rather than humans, and therefore that the metaphysical doctrine of physicalism - that everything must necessarily be a manifestation of something physical - is false.
Notice that there is no mention of AI or Turing machines here. P-zombies only enter the AI debate through additional speculation: If p-zombies are possible, then it is also possible that any machine (Turing or otherwise), no matter how much it might seem to be emulating a human, is at most emulating a p-zombie. As the concept of p-zombies is carefully constructed so as to be beyond scientific examination, such a claim may be impossible to disprove, but it is as vulerable to Occam's razor as is any hypothesis invoking magic or the supernatural.
How do you think a human processes this question?
Quoting A Raybould
Is it possible to get to E = mc^2 without logic?
Quoting A Raybould
What do you mean by that? Do you mean there's a method to insight? Or are insights just lucky guesses - random in nature and thus something computers are fully capable of?
Quoting A Raybould
What's the difference between conceivable and possible?
Quoting TheMadFool
A person who does not just know the answer might begin by asking herself questions like "what does it mean for cheese to melt?" "what causes it to do so?" "what does a refrigerator do?" and come to realize that the key to answering the question posed may be reached through the answers to two subsidiary questions: what is the likely state of the cheese initially, and how is its temperature likely to change after being put into a refrigerator?
At this point, I can imagine you thinking something like "that's just a deductive logic problem", and certainly, if you formalized it as such, any basic solver program would find the answer easily. The part that is difficult for AI, however, is coming up with the problem to be solved in the first place. Judging by the performance of GPT-3, it would likely give good answers to questions like "what causes melting" and "what is a refrigerator?", but it is unable to put it all together to reach an answer to the original question.
It gets more interesting when we consider a slightly more difficult problem: for "cheese", substitute the name of a cheese that the subject has never heard of (there are some candidates here). There is a good chance that she will still come up with the right answer, even if she does not suspect that the object is a form of cheese, by applying suitable general principles and some inductive thinking. Current AI, on the other hand, will likely be flummoxed.
Quoting TheMadFool
That is beside the point. To think that the use of logic in getting to E = mc^2 somehow implies that, once you can get a machine to do logic, there's "nothing special" in getting it to understand things, is, ironically, a failure to understand the role (and limits) of logic in understanding things.
Ultimately, you are arguing against the straightforward empirical fact that current AI has trouble understanding the information it has.
Quoting TheMadFool
Neither of the above. There is a method to solving certain problems in formal logic, that does a breadth-first search through the tree of all possible derivations from the given axioms, but that is nothing like insight: for one thing, there is no semantic content to the formulae themselves. (One of the first successes in AI research, Logic Theorist, proved many of the early theorems from Principia Mathematica, and as doing so is considered a sign of intelligence in people, some thought that AI was close to being a solved problem. They were mistaken.)
What I was thinking is this: if you formalized the whole of classical physics, and started a program such as the above on discovering what it could deduce, the chances that it would come up with E=mc^2 before the world comes to an end are beyond-astronomically small (even more importantly, such a program would not understand the importance of that particular derivation, but that is a separate issue.) The reason for this is the combinatorial complexity of the problem - the sheer number of possible derivations and how fast they grow at each step (even the Boolean satisfiability problem 3-SAT is NP-complete.)
Actually, I have since realized that even this would not be successful in getting to E = mc^2: to get there, Einstein had to break some 'laws' of physics, treat them as approximations, and substitute more accurate alternatives that were still consistent with everything that had been empirically determined. That's not just logic at work.
Lucky guessing has the same problem, and anyone dismissing Einstein's work as a lucky guess just does not understand what he did. There is something more to understanding than any of this, and the fact that we haven't nailed it down yet is precisely the point that I am making on this tangential issue of whether understanding remains a tough problem for AI.
Quoting TheMadFool
Consider the example I gave earlier: I can conceive of the Collatz conjecture being true and of it being false, but only one of these is possible. This situation exists because it is either true or it is false, but so far, no-one has found a proof either way.
In everyday usage, the words might sometimes be considered synonymous, but in the contexts of metaphysics and modal logic, which are the contexts in which the p-zombie argument is made, 'possible' has a specific and distinct meaning. As the example shows, there's an aspect of 'so far as we know' to conceptions, which is supposed to be resolved in moving to metaphysical possibility - when we say, in the context of modal logic, that there is a possible world in which X is true, we are not just saying that we suppose there might be such a possible world. We should either assert it as an axiom or deduce it from our axioms, and if Chalmers had done the former, physicalists would simply have said that the burden was on him to justify that belief (it gets a bit more complicated when we make a counterfactual premise in order to prove by contradiction, but that is not an issue here.)
If the two words were synonymous, Chalmers would not have spent any time in making the distinction and in attempting to get from the former to the latter, and his opponents would not be challenging that step.
Quoting A Raybould
Your ideas are interesting but one thing we agree on is the necessity for logic. Now, the question that pops into my mind is this: what does anyone mean when s/he says something like, "I understand."? To me, understanding is just a semantics game that's structured with syntax. The latter, syntax, doesn't seem to be an issue with computers; in fact computers are veritable grammar nazis. As for the former, semantics, what's the difficulty in associating words to their referents in a computer's memory? That's exactly what humans do too, right? The question for you is this: is there a word with a referent that's impossible to be translated into computer-speak? I would like to know very much, thank you.
Quoting A Raybould
Quoting A Raybould
You talk of "combinatorial complexity" and the way you speak of Einstein's work suggests to me that you think E=mc^2 to be nothing short of a miracle. May I remind you that Newton once said that he achieved what he did only by standing on the shoulder's of giants. There's a long line of illustrious predecessors that paves the way to many scientific discoveries in my opinion. You also seem to ignore serendipity - times when people simply get lucky and make headway with a problem. You seem to be of the view that there's a method behind sheer luck, as if to say there's a way to influence the outcome of a die when in fact it can't be done.
Quoting A Raybould
Just curious here. I feel that I've not understood you as much as I'd have liked but bear with me a while...
If conceivability and possibility are different then the following are possible and I'd like some examples of each:
1. There's something conceivable that's impossible
2. There's something possible that's inconceivable
Quoting TheMadFool
I have no idea what that means. I hope that it means more than "understanding is semantics with syntax", which is, at best, a trite observation that does not explain anything.
Searle says that syntax can not give rise to semantics, and claims this to be the lesson of his "Chinese Room" paper. I don't agree, but I don't see the relationship as simple, either.
Quoting TheMadFool
This is beside the point, as translation does not produce meaning, whether it is into "computer-speak" or anything else. Translation sometimes requires understanding, and it is specifically those cases where current machine translation tends to break down.
You have not, so far, addressed a point that is relevant here: the problem current AI has with basic, common-sense understanding is not in solving logic problems, but in formulating the right problem in the first place.
If you really think you have solved the problem of what it takes to understand something, you should publish (preferably in a peer-refereed journal), as this would be quite a significant advance in the study of the mind. At the very least, perhaps you could address an issue that I have raised twice now: if, as you say, there's nothing special to understanding, and semantics is just associating words to their referents in a computer's memory, how come AI is having a problem with understanding, as is stated in the paper I linked to? Do you think all AI researchers are incompetent?
Quoting TheMadFool
Well I don't, so I think I can skip your arguments against it - though not without skimming them to see if you made any point that stands on its own. In doing so, I see that you put quotes around combinatorial complexity, as if you thought it was beside the point, but it is very much to the point that humans achieve results that would be utterly infeasible if the mind worked like current theorem solving programs.
Quoting TheMadFool
They may be possible, but it is certainly not necessary that there must be something possible that's inconceivable - and if there is, then neither me, you nor anyone else is going to be able to say what it is. On the other hand, in mathematics, there are non-constructive proofs that show something is so without being able to give any examples, and it seems conceivable to me that in some of these cases, no example ever could be found. If this is so, then whether these things should be regarded as inconceivable or not strikes me as a rather subtle semantic issue.
I have twice given you an example of the former: If the Collatz conjecture is true, then that it is false is conceivable (at least until a proof is found) but not possible, and vice-versa. It has to be either one or the other.
By the way, this example is a pretty straightforward combination of syntax, semantics and a little logic, so how do you account for your difficulty in understanding it?
The problem here is simple: what does it mean to understand? Computers are symbol manipulators and that means whatever can be symbolized, is within the reach of a computer. If you believe there's more to understanding than symbol manipulation, you'll have to make the case that that's the case. I didn't find anything in your posts that achieves that. In short, I contend that even human understanding is basic symbol manipulation and so the idea that it's somehow so special that computers can't handle is not true as far as I'm concerned.
Searle's argument doesn't stand up to careful scrutiny for the simple reason that semantics are simply acts of linking words to their referents. Just consider the sentence, "dogs eat meat". The semantic part of this sentence consists of matching the words "dog" with a particular animal, "eat" with an act, and "meat" with flesh, i.e.to their referents and that's it, nothing more, nothing less. Understanding is simply a match-the-following exercise, something a computer can easily accomplish.
Quoting A Raybould
How do you think machine translations work? Each word in a language is matched with another word in a different language. The idea of language translation, as you rightly pointed out, is for humans is an exercise in semantics and yet, despite some mistranslations, computers do a pretty good job. Ask yourself the question: what exactly does understanding semantics mean if a machine, allegedly incapable of semantics, can do as good a job as a human translator of languages?
Quoting A Raybould
While I'm not claiming I'm correct in all what I've said, I've heard that even the very best expert can and do make mistakes.
Quoting A Raybould
Kindly furnish the definitions of "conceivable" and "possible". I'd like to see how they differ, if you don't mind.
Quoting bongo fury
I'll have another rummage.
I expect that you, like "they" in the story, haven't even considered that "referents" might have to be actual things out in the world? Or else how ever did the "linking" seem to you something simple and easily accomplished, by a computer, even??? Weird.
Yes, I did consider that. Referents can be almost anything, from physical objects to abstract concepts. Most physical objects have perception-based qualities to them. Take red wine for example - it's red, has a certain taste, stored in bottles in a cellar, etc. The set of qualities that define what wine is is linked to the word "wine" in the human mind - this is the essence of semantics and computers, in my opinion, are up to the task.
A similar argument can be made for concepts.
Ok...
Quoting TheMadFool
Ah, so after due consideration you decided not. (The referents don't have to be things out in the world.) This was Searle's frustration.
You can be sure you are in the respectable company of his critics at the time. Also of a probable majority of philosophers and linguists throughout history, to be fair.
They don't have to be but they can be, no? I hope you reply soon to this query.
Why? A quick reply isn't usually a thoughtful one. In my case at least. Actually, I think the site should instigate a minimum time between replies, as well as a word limit.
Quoting TheMadFool
I don't think I've been understood, here. (Take more time?) I was trying to explain why @A Raybould was non-plussed by your statements about semantics. See also @InPitzotl's recent posts here.
It's about
Quoting TheMadFool
They (and I) mean things out there, you mean just more words/data.
"Within the reach" avoids precision where precision is needed. What do you mean, here?
Quoting TheMadFool
Whether it is symbol manipulation is beside the point. What's at issue here is my statement that "[Turing-like] tests could be further improved by focusing on what a machine understands, rather than what it knows" and your reply that you don't see anything special in understanding. Being symbol manipulation does not automatically make it simple, and your explanations, which invoke symbol manipulation whithout showing what sort of manipulation, are just part of the reason for thinking that it is not.
Quoting TheMadFool
That is a view I have not stated and do not hold. My position has consistenly been that understanding is not a simple issue and that it remains a significant obstacle for AI to overcome. I have also taken the position that current logic solvers are not sufficient to give a machine the ability to understand the world, which should not be mistaken for a claim that no form of symbol manipulation could work. To be clear, my position on consciousness is that I suppose that a digital computer could simulate a brain, and if it did, it would have a mind like that of the brain being simulated.
Quoting TheMadFool
Please expand on 'match-the-following', as I cannot imagine any interpretation of that phrase that would lead to a computer being able to understand anything to the point where it would perform reasonably well on "common-sense physics" problems (in fact, perhaps you could work through the "if you put cheese in a refrigerator, will it melt?" example?)
Quoting TheMadFool
You have taken hold of the wrong end of the stick here. I was replying to your question "Is there a word with a referent that's impossible to be translated into computer-speak?" by pointing out that it is irrelevant to the issue because translation does not create or add meaning. In turn, it is also irrelevant to this point whether the translation is done by humans or machines: neither of them create or add meaning, which is delivered in the original text.
Quoting TheMadFool
You are barking up the wrong tree here, precisely because translation does not modify the semantics of its input. To address the matter at hand, you would need an example that does demand understanding. We have that, at least to a small extent, in the common-sense physics questions discussed in the paper I linked to, and even here the performance of current AI is weak (note that this is not my assessment or that of critics; it is from the team which developed the program.) You have avoided addressing this empirical evidence against your claim that machine understanding is simple, until...
Quoting TheMadFool
Really? Do you understand that, for this excuse to work, in would not take just one or two experts to make a few mistakes; it would require the entire community to be mistaken all the time, never seeing the simple solution that you claim to have but have not yet explained?
Quoting TheMadFool
At last! Back to the main issue. I will start by quoting what I wrote previously:
Quoting A Raybould
To expand on that, one could hold that any sentence in propositional form is conceivable, as it is conceived of merely by being expressed (some people might exclude propositions that are false a priori, but a difficulty with that is that we don't always (or even often) know whether that is the case.)
In the context of modal arguments, of which the p-zombie argument is one, for the sentence to be possible, it must be true in a possible world. In modal logic, if you want a claim that something is possible to be accepted by other people, you either have to get them to accept it as an axiom, or you must derive it from axioms they have accepted.
I am not sure if the above is going to help, because the debate over whether Chalmers can go from conceivability to possibility is, in part, a debate over what, exactly, people have accepted when they accept the conceivability of p-zombies. What seems clear, however, is that neither side is prepared to say that they are the same.
By the way, your positions seem to generally physicalist, except that you are troubled by p-zombies, which are intended to be anti-physicalist. AFAIK, this is quite an unusual combination of views.
Whether they are things out in the world, or merely more words referring to those things.
Quoting TheMadFool
Yep. So what is it that a computer so easily (according to you) links to the word "water"? The referent you just described, or merely the description?
The description consists of referents.
While the initial association of symbols may require human input, once the work is complete, a computer can use the databank just as humans do.
Quoting A Raybould
You insist that human understanding is not something a computer can do but what's the argument that backs this claim? I'd like to see an argument if it's all the same to you.
Quoting A Raybould
What I'm saying is very simple. Semantics is nothing more than associating words with something else - be it a concrete as a stone or as abstract as calculus. Associating two things is easily done by a computer and ergo, in my humble opinion, semantics can be handled by a computer.
To back up my position, I'll ask a few questions:
1. When you read the word "water" and understand it, what's going on in your head?
By the way, you haven't furnished the definitions for "conceivable" and "possible".
Ok, well to see "why people make such a big deal of understanding" you need to see that they are interested in how we link the word "water" to the water itself, and not merely to more words for water.
"Referent" usually refers to the designated object itself, not to other words, semantically related or not.
How do we do it, link the word "water" to the water itself, in your opinion?
By learning to agree (or disagree) with other people that particular tokens of the word are pointed at particular instances of the object.
That's to say there is no meaning except in the sense of a consensus. What makes you think computers can't do that? Can't one computer use the same word-referent associations as another?
If you like. Is that an objection?
Quoting TheMadFool
What, agree and disagree about where each other's words have 'landed', out in the world? If by computers you mean some future AI, then sure. This would no doubt be a few steps more advanced than, say, being able to predict where each other's ball has (actually) landed. Which I assume is challenging enough for current robots.
I am replacing my original reply because I do not think the nitpicking style that this conversation has fallen into is helpful.
From your explanations and 'water' question in your latest reply, it seems increasingly clear to me that we have very different ideas of what understanding is. For you, it seems to be something such that, if a person memorized a dictionary, they would understand everything that is defined in it. For me, it is partly an ability to find the significant, implicit connections between the things you know, and there is also a counterfactual aspect to it: seeing the consequences if things were different, and seeing what needs to change in order to get a desired result.
Given these different conceptions, it is not surprising that you might think it is an easy problem, while I see significant difficulties. I will not repeat those difficulties here, as I have already covered them in previous posts.
As for what I believe, no extant computer(+program) can perform human-like understanding, but I expect some future computer could do so.
With regard to conceivability versus possibility, I gave my working definitions in my previous post, though not spelled out in 'dictionary style.' For completeness, here are the stripped-down versions:
Conceivable: Anything that can be stated as a proposition is conceivable.
Possible: In the context of modal logic, which is the context of Chalmers' argument, something is possible if and only if it can be stated as a proposition that is true in some possible world.
I do not think reducing them to bare definitions is very helpful, and by doing so, perhaps I can persuade you of that. I urge you to take another look at the Collatz conjecture example from before.
I agree. For your kind information, I suspect my idea of understanding is simpler than yours; that probably explains why I feel computers are capable of it.
What exactly does understanding mean for you?
Let me illustrate what understanding means (for me):
1. "Trees need water" consists of three words viz. "trees", "need" "water". It's clear that all three of them have referents. "Trees" and "water" being concrete objects, they maybe easy for a computer to make the connection between these words and their referents. The referent for "water" will consist of the sensory data that are associated with water and to that we can add scientific knowledge that water is H20, has Hydrogen bonds, etc.
The same can be done with "trees". To cut to the chase, understanding the words "trees" and "water" is simply a process of connecting a specific set of sensory and mental data to these words.
Coming to the word "need" we immediately recognize that the word refers to a concept i.e. the word has an abstract referent. The concept of need is a pattern abstracted from instances such as the relationship between plants and water, animals and oxygen, fire and heat, cars and gasoline, etc. In other words, if computers can be programmed to seek patterns the way humans can then, even abstract concepts, ergo abstract referents, are both within the reach and grasp of computers so programmed.
In summary, understanding is about 1) associating sets of [sensory & mental] data (sensory as in through the sense organs and mental as in knowledge within a given paradigm) and 2) pattern detection.
Which of the two aspects of understanding delineated above are difficult or impossible for a computer in your opinion?
What's your definition of understanding?
Quoting A Raybould
I covered this above.
Quoting A Raybould
Thanks but I have an issue with your definition of "conceivable". According to you, unlike possibility, conceivability has no logical significance at all. When I say, "p-zombies are conceivable" I'm making the claim about p-zombies. That I can say that sentence ("can be stated") is the least of my concerns. Your definition of "conceivable" falls short of being relevant to the issue of whether p-zombies are possible/conceivable. Thanks anyway.
Do you mean that human understanding is reducible to computer logic but that we haven't the technology to make it work? If yes then that means you agree with me in principle that human understanding isn't something special, something that can't be handled by logic gates inside computers.
Only in the almost trivial sense that neurons are quite evidently some kind of switch or trigger.
Quoting TheMadFool
I roughly agree with you now (maybe, or maybe the switches will have to be actual neurons; we don't yet know), since you're talking about way off in the future.
But do you at last see the trouble here,
Quoting TheMadFool
?
The problem with Searle's Chinese room is that the man in the room does understand something - the rules he is following - write this symbol when you see this symbol. It's just not the same rules that Chinese speaking people follow when using those same symbols.
The meaning of symbols can be arbitrary. Just look at all the different words from different languages that refer to the same thing. When we aren't using the same rules for the same symbols it can appear as if one of us isn't understanding the symbols.
That's what understanding is - having a set of rules for interpretting symbols. In this sense, computers understand thanks to their programming (a set of rules).
What would a non-trivial sense look like? I mean your beliefs on human understanding seems to me as something approaching, oddly and also intriguingly, incomprehensibility for/by humans themselves. Unless, ofcourse, you have something to say about it...what's your take on understanding?
Quoting bongo fury
To make it short, no.
You're right, will do as the inclination strikes.
Quoting A Raybould
Oh my. That's not true. But first for the record let me say that I agree with you. A TM could perhaps convincingly emulate but never implement an actual human mind. I don't believe the mind is a TM.
But many smart people disagree. You have all the deep thinkers who believe the entire universe is a "simulation," by which they mean a program running on some kind of big computer in the sky. (Why do these hip speculations always sound so much like ancient religion?) We have many people these days talking about how AI will achieve consciousness and that specifically, the human mind IS a TM. I happen to believe they're all wrong, but many hold that opinion these days. Truth is nobody knows for sure.
I've read many arguments saying that minds (and even the entire universe) are TMs. Computations. I don't agree, but I can't pretend all these learned opinions aren't out there. Bostrom and all these other likeminded characters. By the way I think Bostrom was originally trolling and is probably surprised that so many people are taking his idea seriously.
Quoting A Raybould
I'm perfectly willing to stipulate that a p-zombie is physically indistinguishable. I made the assumption, which might be wrong, that their impetus or mechanism of action is driven by a computation. That is, they're totally human-like androids run by a computer program.
If you are saying the idea is that they're totally lifelike and they have behavior but the behavior is not programmed ... then I must say I don't understand how such a thing could exist, even in a thought experiment. Maybe I should go read Chalmers.
Quoting A Raybould
You and I would both like to believe that. But neither of us has evidence that the mind is not a TM, nor do we have hours in the day to fight off the legion of contemporary intellectuals arguing that it is.
Roger Penrose is famous for arguing that the mind is not a computation, and that it has something to do with Gödel's incompleteness theorem being solved in the microtubules. Nobody takes the idea seriously except as a point of interest. Sir Roger's bad ideas are better than most people's good ones.
But you can't be unaware that many smart people argue that the mind is a TM.
We don't know of anything that works in a "non-Turing, non digital way." There are mathematical models of hypercomputation or supercomputation in which one assumes capabilities beyond TMs. But there's no physics to go along with it. Nobody's ever seen hypercomputation in the physical world and the burden would be on you (and I) to demonstrate such.
Quoting A Raybould
I recall this argument. It's wrong. If our minds are a logical or necessary consequence of our physical configuration, and a p-zombie is identical to a human, then the p-zombie must be self-aware.
Otherwise there is some "secret sauce" that implements consciousness; something that goes beyond the physical. Some argument for physicalism. You just refuted it. Maybe I"m misunderstanding the argument. But if the mind is physical and a p-zombie is physically identical, then a p-zombie has a mind. If a p-zombie is physically identical yet has no mind, then mind is NOT physical. Isn't that right?
Quoting A Raybould
Only, in my opinion, because not every philosopher understands the theory of computation.
What animates the p-zombie? If it's a mindless machine, it must have either a program, or it must have some noncomputable secret sauce. If the latter, the discovery of such a mechanism would be the greatest scientific revolution of all time. If Chalmers is ignoring this, I can't speak for him nor am I qualified to comment on his work.
Quoting A Raybould
As I mentioned earlier, it's entirely possible that my next door neighbor is only emulating a human. We can never have first-hand knowledge of anyone else's subjective mental states.
I still want to understand what is claimed to animate a p-zombie. Is it a computation? Or something extra-computational? And if it's the latter, physics knows of no such mechanism.
Quoting A Raybould
I think you've motivated me to at least go see what Chalmers has to say on the subject. Maybe I'll learn something. I can see that there must be more to the p-zombie argument than I'm aware of.
ps -- I just started reading and came to this: "It seems that if zombies really are possible, then physicalism is false and some kind of dualism is true."
https://plato.stanford.edu/entries/zombies/
This tells me that my thinking is on the right track. If we are physical and p-zombies are physically identical then p-zombies are self-aware. If they aren't, then humans must have some quality or secret sauce that is non-physical.
I would raise an intermediate possibility. The mechanism of mind might be physical but not computable. So we have three levels, not two as posited by the p-zombie theorists:
* Mind is entirely physical.
* Mind is entirely physical but not necessarily computable, in the sense of Turing 1936. It might be some sort of hypercomputation as is studied by theorists.
* Mind might be non-physical. In which case we are in the realm of spirits and very little can be said.
pps --
AHA!!
" Proponents of the argument, such as philosopher David Chalmers, argue that since a zombie is defined as physiologically indistinguishable from human beings, even its logical possibility would be a sound refutation of physicalism, because it would establish the existence of conscious experience as a further fact."
https://en.wikipedia.org/wiki/Philosophical_zombie
This is exactly what I'm understanding. And I agree that I probably made things too complicated by inserting computability in there. The p-zombie argument is actually much simpler.
Well that counts as research for me these days. A couple of paragraphs of SEP and a quote-mine from Wiki. Such is the state of contemporary epistemology. If it's on Twitter it's true.
There is no point in discussing your own private definition of 'understanding' - no-one can seriously doubt that computers are capable of performing dictionary look-up or navigate a predefined network of lexical relationships; even current AI can do much more than just that.
We can make no judgment, however, of whether AI is performing at human-like levels by only looking at simple examples, and the fact remains that AI currently has problems with certain more demanding cognitive tasks, such as with "common-sense physics" (as I mentioned previously, that is not just my opinion, it is a quote from those who are actually doing the work.) You have given no plausible explanation for how your concept of understanding, and of how it can easily be achieved, solves this problem, and in your only attempt to explain away why, if it is so easy, it remains an acknowledged problem in actual AI research, you implied that the whole AI community has consistently failed to see what is obvious to you.
There is no reasonable doubt that AI currently has a problem with something here; I just don't know what it is called in your personal lexicon.
Personal lexicons come up again in the issue of 'conceivable' vs. 'possible', where the definition I attempted of 'conceivable' apparently doesn't match yours. There is no point in getting into a "you say, I say" argument, but we don't have to: it is a straightforward fact that the distinction between 'conceivable' and 'possible' is widely accepted among philosophers and is central to Chalmer's p-zombie argument. I will grant that it is conceivable, and even possible, that you are right and they are all wrong, but I don't think it is probable.
You would be more convincing if you could explain where the example I gave earlier, using the current status of the Collatz conjecture, goes wrong.
Even in Bostrom's simulation argument, neither brains nor minds are TMs: in that argument, I (or, rather, what I perceive as myself) is a process (a computation being performed), and what I perceive as being the rest of you is just data in that process. To confuse a process (in either the computational sense here, or more generally) with the medium performing the process is like saying "a flight to Miami is an airplane." A computation is distinct from the entity doing the computation (even if the latter is a simulation - i.e. is itself a computation - they are different computations (and even when a computation is a simulation of itself, they proceed at different rates in unending recursion.))
I recognize that this loqution is fairly common - for example, we find Searle writing "The question is, 'Is the brain a digital computer?' And for the purposes of this discussion I am taking that question as equivalent to 'Are brain processes computational?" - but, as this quote clearly shows, this is just a manner of speaking, and IMHO it is best avoided, as it tends to lead to confusion (as demonstrated in this thread) and can prime the mind to overlook certain issues in the underlying question (for example, if you assume that the brain is a TM, it is unlikely that you will see what Chalmers is trying to say about p-zombies.) To me, Searle's first version of his question is little more than what we now call click-bait.
What's the public definition of understanding. Since no definition was agreed upon I thought I might just throw in my own understanding of understanding. What's your definition, if I may ask?
Quoting A Raybould
I don't know what's so uncomputable about understanding. As far as I can tell, all that's required are:
1. Word-referent connection
2. Pattern recognition
Is there anything else to understanding? If there is I'd like to know. Thanks.
Quoting A Raybould
What is the difference between the two? If they're the same then it's true that conceivable if and only if possible.
If they're different then it's possible that conceivable but not possible and possible but not conceivable. Please provide examples of both scenarios for my benefit. Thanks.
If the word simulation means something other than computation, you need to state clearly what that is; and it has to be consistent either with known physics; or else stated as speculative physics.
I'll agree that Bostrom and other philosophers (Searle included) don't appear to know enough computability theory to realize that when they say simulation they mean computation; and that when they say computation they must mean a TM or a practical implementation of a TM. If not, then what?
When we simulate gravity or the weather or a first person shoot-'em-up video game or Wolfram's cellular automata or any other simulation, it's always a computer simulation. What other kind is there?
And when we say computation, the word has a specific scientific meaning laid out by Turing in 1936 and still the reigning and undefeated champion.
Now for the record there are theories of:
* Supercomputation; in which infinitely many instructions or operations can be carried out in finite time; and
* Hypercompuation; in which we start with a TM and adjoin one or more oracles to solve previously uncomputable problems.
Both supercomputation and hypercomputation are studied by theorists; but neither are consistent with known physical theory. The burden is on you to be clear on what you mean by simulation and computation if you don't mean a TM.
Quoting A Raybould
But what do you mean by computation? Turing defined what a computation is. If you mean to use Turing's definition, then you have no disagreement with me. And if you mean something else, then you need to clearly state what that something else is; since the definition of computation has not changed since Turing's definition.
Quoting A Raybould
I have done no such thing. I don't know why you'd think I did. A computation is not defined by the medium in which it's implemented; and in fact a computation is independent of its mode of execution. I genuinely question why you think I said otherwise.
If you agree that you and I are "processes," a term you haven't defined but which has a well-known meaning in computer science with which I'm highly familiar, then a process is a computation. You can execute Euclid's algorithm with pencil and paper or on a supercomputer, it makes no difference. It's the same computation.
Quoting A Raybould
You're arguing with yourself here. I have never said anything to the contrary. A computation is independent of the means of its execution. What does that have to do with anything we're talking about?
Quoting A Raybould
Searle also, in his famous Chinese room argument, doesn't talk about computations in the technical sense; but his argument can be perfectly well adapted. Searle's symbol lookups can be done by a TM.
And again, so what? You claim the word simulation doesn't mean computation; and that computation isn't a TM. That's two claims at odds with reality and known physics and computer science. The burden is on you to provide clarity. You're going on about a topic I never mentioned and a claim I never made.
Quoting A Raybould
But a computation is a very specific technical thing. If I start going on about quarks and I say something that shows that I'm ignorant of physics and I excuse myself by saying, "Oh that was just a manner of speaking," you would label me a bullshitter.
If you mean to use the word computation, you have to either accept its standard technical definition; or clearly say you mean something else, and then say exactly what that something else is.
Quoting A Raybould
I'm not confused. My thinking and knowledge are perfectly clear. A computation is defined as in computer science. And if you mean that we are a "simulation" in some sense OTHER than a computation, you have to say what you mean by that, and you have to make sure that your new definition is compatible with known physics.
Quoting A Raybould
I understand exactly what Chalmers is saying about p-zombies now that I re-acquainted myself with the topic as a result of this thread.
But you're going off in directions.
What do you mean by simulation, if not a computer simulation? And what do you mean by a computation, if not a TM?
Quoting A Raybould
Whatever. I'm not Searle and he got himself into some #MeToo trouble and is no longer teaching. Why don't you try talking to me instead of throwing rocks at Searle?
What part of 'a computation is what a Turing machine does, not what it is' do you not understand? At least until we sort that out, I am not going to read any more of this jumble.
Best you don't, since I couldn't have been more clear.
A TM is not a physical device. It's an abstract mathematical construction. A computation, by definition, is anything that a TM can do. This isn't me saying this, it's Turing followed by 80 years worth of computer science saying that.
If you think there's something that counts as a computation, that
a) Can not be implemented by a TM; and
b) Is consistent with known physics;
then by all means tell me what that is to you.
I rather suspect that's true, unfortunately.
I haven't seen your handle much before. People who know me on this board know that I'm perfectly capable of getting into the mud. I'm sorely tempted at this moment but will resist the urge. I say to you again:
A TM is not a physical device. It's an abstract mathematical construction. A computation, by definition, is anything that a TM can do. This isn't me saying this, it's Turing followed by 80 years worth of computer science saying that.
If you think there's something that counts as a computation, that
a) Can not be implemented by a TM; and
b) Is consistent with known physics;
then by all means tell me what that is to you.
Quoting fishfry
I will treat that comment with all the respect it deserves.
Quoting fishfry
Regardless, the question I asked a couple of posts ago applies either way.
Quoting fishfry
...but it is not the computation that the abstract machine is computing. I covered that in yesterday's post.
Do you think that in Bostrom's simulated universes, it's TMs all the way down? Clearly not, as his premises don't work in such a scenario - there's a physical grounding to whatever stack of simulations he is envisioning.
I discussed this at length. You chose not to engage with my questions, my points, or my arguments. You failed to demonstrate basic understanding of the technical terms you're throwing around. You repeatedly failed to define your terms "process" and "simulation" even after my repeated requests to do so.
This is no longer productive.
First, let me make one thing clear (once again): The issue is not whether understanding is uncomputable, and if you think I have said so, you are either misunderstanding something I wrote, or drawing an unwarranted conclusion. The issue here is your insistence that there is nothing special about understanding and that it is a simple problem for AI.
I have already given you a working definition that you chose to ignore. Ignoring me is one thing, but if, instead, you were to look at what real philosophers are thinking about the matter, you would see that, though it is a work in progress, at least one thing is clear: there is much more to it than you suppose.
We can, however, discuss this matter in a way that does not depend on a precise definition. If, as you say, having AIs understand things is simple, then how come the creators of one of the most advanced AI programs currently written acknowledge that understanding common-sense physics, for one thing, is still a problem? Here we have a simple empirical fact that really needs to be explained away before we can accept that understanding (regardless of how you choose to define it) actually is simple - yet many posts have gone by without you doing so.
Of course, anyone reading your 'explanation' of how to do machine understanding will have a problem implementing it, because it is so utterly vague. It most reminds me of many of the dismissive posts and letters-to-the-editor written after IBM's Watson's success in the Jeopardy contest: "it's just database lookup" was typical of comments by ignoramuses who had no idea of how it worked and by how much it transcended "just" looking up things in a database.
Quoting TheMadFool
When I read this, I got the distinct feeling that I was dealing with a bot, which would be quite embarrassing for me, given the original topic of this thread! Things that tend to give away a bot include blatant non-sequiturs, a lack of substance, a tendency to lose the thread, and repetition of errors. You asked essentially the same question as this one here (complete with the same basic error in propositional logic) a few posts back, but when I provided just such an example (the same one as I had given more than once before) you ignored it and went off in a different direction, only to return to the same question now.
I am tempted to just quote my reply from then, but I will spell it out more formally, so you can reference the first part you don't agree with:
In other words, it's implied, you feel understanding is uncomputable i.e. there is "something special" about it and for that reason is beyond a computer's ability.
Quoting A Raybould
I'll make this as clear as I possibly can. Do you think humans are different from machines in a way that gives humans certain abilities that are not replicable in machines? Do we, humans, not obey the laws of chemistry and physics when we're engaged in thinking and understanding? It seems to me that all our abilities, including understanding, arise from, are determined by, chemical and physical laws all matter and energy must obey. The point being there's nothing magical going on in thinking/understanding - it's just a bunch of chemical and physical processes. We're, all said and done, meat machines or wet computers.
There's nothing physically or chemically impossible going on inside our heads. Hence, I maintain that thinking/understanding is, for sure, computable.
Quoting A Raybould
obscurum per obscurius
Define the words "conceivable" and "possible" like a dictionary does.
Absolutely not. As you are all for rigor where you think it helps your case, show us your argument from "there's something special about understanding" to "understanding is uncomputable."
Quoting TheMadFool
As I have pointed out multiple times, that is not the issue in question. Here, you are just making another attempt to change the subject, perhaps because you have belatedly realised that you cannot sustain your original position? Until you have completed the above task, stop attempting to attribute to me straw-man views that I do not hold and have not advocated.
Quoting TheMadFool
Now you are just trolling, and using Latin does not alter that fact. Here we have a straightforward argument that you apparently don't agree with, but for which you cannot find a response.
Quoting TheMadFool
I see you are reverting to bot-like behavior, as outlined in my previous post. We have been round this loop before. I see, from other conversations, that you frequently use demands for definitions to browbeat other people when you have run out of arguments, to take the discussion in a different direction... Well, it won't work here: I am not going to follow you in another run around the rabbit-warren until you have addressed the specific argument here.
1. Connect symbols to referents (computable). Even Google search can do that. For instance when I type "water" into the search bar, Google takes me to articles that contain the word "water" and even images of glasses of water.
2. Recognize patterns (computable). Google search couldn't have taken me to articles on water and images of water without it being able to identify patterns.
As for the definitions of "conceivable" and "possible", I'd like to see them in a familiar format please, like in a dictionary.
One possible conclusion from this equation is that p-zombies, as defined, cannot exist.
:chin:
I think it's obvious that they cannot possibly exist. Mind you, they don't actually exist. They are just a mind experiment designed to probe the nature of consciousness.
I have noticed a pattern here: you will post a claim, I will respond, then you will raise a different issue as if you had no counter-argument. A post or two later, however, the first issue will rise again, zombie-like, as if it had never been discussed before.
For convenience, I will list these recurring arguments and my responses. That way, you can make a comment that merely states the number of the argument du jour, and I can reply by picking a corresponding response ID and just posting that. It will make things so much easier!
A1: Understanding is simple, because understanding is computable.
R1: Being computable does not necessarily entail simplicity. If this were the case, the whole of AI would be simple, the ABC conjecture would have a simple proof (if there really is one), etc.
A2: Understanding is simple, because understanding is just a matter of connecting symbols to referents and recognizing patterns.
R2.1 This is too vague to establish simplicity. It is so vague that you could make the same claim for any aspect of AI, or AI as a whole (like the people who dismiss current AI as "just database lookup"), but if it is that simple, how come there are still outstanding problems?
R2.2: Regardless of what definition you put forward, the claim that it is simple to implement is inconsistent with the fact that current AI has ongoing difficulties with, for example, understanding common-sense physics.
A3: Understanding is simple, as shown by this simple example of what I understand from the word 'water'.
R3.1: You cannot establish simplicity through only simple examples, unless there are only simple examples. As it happens, there are difficult examples here, such as the afforementioned difficulty with understanding common-sense physics.
R3.2: If you could establish simplicity through simple examples, then the whole of mathematics would be simple, as established by the fact that you can teach five-year-olds to add two numbers.
Did I miss any? Your help in completing the list would be appreciated!
So, according to this list, your latest post is A2 A3, to which I reply R2.2 R3.1 (I have some flexibility here.)
Quoting TheMadFool
We have already been around this define / whats-the-difference? / define loop once before, and as I made clear, I have no intention of going round again until you address the whats-the-difference example.
If you want to get into definitions, it's your turn to offer some, so how about if you proffer definitions which make the Collatz conjecture example/argument invalid or unsound? (Or, if you find that infeasible, you could simply say which premise or conclusion you first disagree with, and we can proceed from there.)
You've failed to notice your own pattern.
Quoting A Raybould
I never claimed understanding as simple. I said it's computable.
Quoting A Raybould
I never brought up the notion of simplicity. I don't know where you got that from. Also, 1)word-referent matching and 2) pattern recognition, as far as I can tell, aren't vague at all. If you insist they are, demonstrate them to be so.
Quoting A Raybould
I don't see a necessary connection between conceivable & possible and the Collatz conjecture. Are you implying the meanings of conceivable and possible are based off of the Collatz conjecture? If yes, I'm more than happy to see a proof of that.
Quoting TheMadFool
Quoting TheMadFool
Quoting TheMadFool
...and so on. These are not 'gotcha' quotes taken out of context; the alleged simplicity of understanding is a big part of your claim that there is nothing special about it.
Quoting TheMadFool
But the issue is not whether it is computable, as I have repeatedly had to remind you. Do you not remember this?
Quoting A Raybould
I am quite willing to believe that initially, you may have merely misunderstood what I meant, by drawing an unjustified conclusion such as the one above, but to continue as if this is the issue in contention, after having been repeatedly corrected on the matter, is another example of trollish behavior, and I will continue to call you on it wherever I see it, whether it is in response to me or someone else (in fact, if it were not for this aspect of your replies, I would drop the issue as being merely a misunderstanding and a difference of opinion.)
Quoting TheMadFool
For one thing, they are vague, when considered as an explanation of understanding, in that they lack the specificity needed for it to be clear that anything having just those two capabilities would necessarily understand, say, common-sense physics or Winograd schema. I am willing to believe that a machine capable of understanding these things could be described as having these capabilities, but I am also pretty sure that many machines, including extant AIs such as GPT-3, could also be so described, while lacking this understanding. If so, then this description lacks the specificity to explain the difference between machines that could and those that cannot understand these things.
Quoting TheMadFool
No - I should have made it clear that the Collatz conjecture is just something for which neither a proof nor a refutation has been found so far; any other formal conjecture would do as well in its place. The essence is that there are two conceivable things here, and we know that only one of them is possible (even though we don't know which one), so the other (whichever one it is) is conceivable but not possible.
Firstly, I admit that I used the word "simple" in reference to understanding but only to characterize its nature as not something magical and beyond the scope of computers. Understanding, as it appears to be, is probably a complex phenomena, nevertheless computable. That's what I mean when I said "I don't know why people make such a big deal of understanding - it's very simple. Very simple in the sense of being reducible to logic, something computers are capable of. To clarify further, take aerodynamics - the science of flight - and you'll notice a simplicity in the fact that flying involves only 4 forces viz. lift, weight, thrust and drag but building a plane is a complex affair. Similarly, the simplicity in understanding lies in it involving only a couple of actions viz. word-referent matching and pattern recognition but the complexity lies in how these simple actions can be programmed at the required level in a computer.
Quoting A Raybould
I wouldn't go so far as to say the two conditions for understanding I mentioned in my post are complete. Some other conditions may need to be added.
Quoting A Raybould
So, you're saying, with reference to the Collatz conjecture, that it's conceivable for the Collatz conjecture to be true AND false. Put differently, you're claiming:
1. It is conceivable that the Collatz Conjecture is true AND it is conceivable that the Collatz Conjecture is false
That's a contradiction.
Computers as we know them are not aware of the world around them, and that means they cannot realy understand anything, because they don't know that there exists referents out there for words like "trees" or "water".
Let me take your own example to illustrate the point. Here is how, as a human being, I understand the proposition "trees need water": it means to me "IN ORDER TO STAY ALIVE, trees need TO ABSORB SOME MINIMUM AMOUNT OF water PER UNIT OF TIME".
STAY ALIVE: Evidently, dead trees don't need water for anything. The need is related to life and its maintenance.
ABSORB: Evidently, trees don't need water that they can't absorb. They absorb water usually though their root system, so if you just place a glass of water standing next to a tree, you're not providing for its need.
SOME AMOUNT... PER UNIT OF TIME: Evidently their water needs are not infinite. If you don't water to a tree for a day or two, it will be fine. And if you place a tree under water (or water-log its root system) it may well die. So they need SOME water, and some trees require more water than others.
So your seemingly simple proposition, "trees need water" cannot be properly understood by a machine who has no clue about trees and their biology.
I think we have reached the point that we can agree to differ over whether or not there is something special about understanding, because we are approaching the question from different perspectives.
With regard to conceivability: It would be a contradiction to say that 'the Collatz Conjecture is true AND the Collatz Conjecture is false', but 'It is conceivable that the Collatz Conjecture is true AND it is conceivable that the Collatz Conjecture is false' is not the same, at least formally.
To see this, consider sentences of the form 'It is P that the Collatz Conjecture is true AND it is P that the Collatz Conjecture is false'. Substituting 'true' for P leads to a contradiction, but substituting 'uncertain' does not. Without more precise semantics for 'conceivable', we cannot say that we get a contradiction when we substitute it for P (in the Collatz conjecture argument, I avoided giving a full definition of 'conceivable' by saying that everything that has been conceived of is a subset of everything conceivable. You can read 'conceived of' as 'thought of'.)
Chalmer's p-zombie argument is entirely dependent on taking the step from p-zombies being conceivable to being possible, so what he intends these two words to mean, and the relationship between them, is of critical importance. If they have the same meaning, then he is simply asserting that p-zombies are possible, without offering any argument for that claim; to put it another way, he would merely be inviting us to share his belief, without there being any risk of us falling into a contradiction if we decline to do so.
Chalmers, therefore, is walking a narrow path: his definition of 'conceivable' has to be distinct from 'possible', but not so distinct that he needs additional assumptions to get from the former to the latter, and especially not any contingent assumptions, which could be false as a matter of fact.
I know one thing you are thinking right now: So what is Chalmers' definition of 'conceivable'? I am not certain, and I don't think there is an easy answer; the first place to look would be his paper Does Conceivability Entail Possibility?, though it is not an easy read. For what it is worth, my impression of the paper was that he only says it does so in those cases where there are other, independent, reasons for saying that the conceivable thing is possible - which amounts to saying "no, conceivability by itself does not entail possibility", and therefore his 'argument' for the possibility of p-zombies is merely an unargued-for belief.
Thanks for saying that. It is easy to get carried away whan defending a point of view. I do, and I used to do it a lot more; I have to make a conscious effort to back away.
I had some more thoughts on conceivabilty vs. possibility. Most philosophers accept possible-world semantics for dealing with questions of possibility and necessity, in which to say something is possible is to assert that there is a possible world in which it is true, regardless of whether it is true in the actual world (that looks somewhat self-referential, but logicians seem to agree its OK.)
It is also generally held that mathematical truths are necessary truths, and necessary a priori at that. A mathematical fact is true in all possible worlds, and always has been.
Putting these two things together gets tricky when we consider a mathematical conjecture. Given that we do not know whether it is true, we might want to say it is possible that it is true; at another time, we might want to say that it is possible that it is false. Under possible-world semantics, however, if mathematical truths are necessary truths, then one or the other of those statements must be false: if the conjecture is true, it is necessarily true, so there are no possible worlds in which it is false, and vice-versa.
What we want is a way of saying that something might be true, without invoking all the implications that come with possible-world semantics. Saying that it is is conceivable is a way of doing that. (Note that even though, in everyday usage, 'might be true' and 'possibly true' usually mean more-or-less the same thing, they are different when 'possible' is being used in the context of possible-world semantics.)
Chalmers also says p-zombies are 'logically possible', which looks like a strong statement, but it really just says that he is unaware of any facts that could disprove them. Given that he has defined p-zombies in a way that makes them immune from being ruled out by any sort of scientific investigation or discovery whatsoever, this is not saying much.
What about the following statements then:
1. It is possible that a mathematical conjecture is true or false.
Nevertheless,
2. It's impossible that a mathematical conjecture is both true AND false
Compare statements 1 and 2 with:
3. It is conceivable that a mathematical conjecture is true or false
Nevertheless
4. It is jnconceivable that a mathematical conjecture is both true AND false
As you can see, the notion of possibility does have, within it, the idea of all available truth values for a given proposition, just like conceivability does. Most importantly, contradictions are both impossible AND inconceivable which seems to suggest that conceivability supervenes, if that's the right word, on possibility.
But isn't the real question here, that in some certain cases there is no way for us to prove / compute if a conjecture is either true or false? Wasn't this what Turing showed in the first place?
That mathematics, to be logical, has to have unprovable statements, which still are either true or false.
And hence, it would be perhaps provable that consciousness is unprovable.
Yes, I think you are on the right track here, though it may be leading in a surprising direction.
Firstly, there is no difficulty with 'it is possible/conceivable that the Collatz conjecture is true (X)OR it is possible/conceivable that the Collatz conjecture is false', which is the correct way to express the fact of our incomplete knowledge (at least if it is decidable.)
Secondly, your assertion that 'it is inconceivable that a mathematical conjecture is both true AND false' depends on whether false statements are conceivable. You may find it inconceivable that they are, but others may logically disagree.
There is an alternative view here that is... conceivable? It says that it is conceivable that a mathematical conjecture is both true AND false, it is just that we can immediately refute it (i.e. immediately see that it is not possible.) In this view, we had to conceive of it first (to form the thought in our minds), in order to prove that it is not possible.
You might wonder if there is anything that is not conceivable in this view. For one thing, I can conceive of there being inconceivable ideas by virtue of them requiring too much physical information for a brain to contain (and, as information can grow exponentially with the medium in which it is expressed, I do not think this can be avoided by positing an AI larger than a brain.) Maybe the true Theory of Everything is like this. Also, as Eliezer Yudkowsky pointed out, two millennia of philosophizing over epistemology and metaphysics never conceived of the sort of non-local reality that is the only sort allowed by Bell's Inequality; it only became conceivable in the light of new knowledge.
In my previous post, I wrote "Given that we do not know whether [the Collatz conjecture] is true, we might want to say it is possible that it is true; at another time, we might want to say that it is possible that it is false." I have emphasized "at another time" here because in my first draft, I wrote 'and' instead, but changed it, as the original statement would be begging the question I was trying to address. The point I wanted to make in that post is that, under possible-world semantics, and regardless of any difficulties with conjunctions, one cannot even know[1] that 'it is possible that the conjecture is true', because if it is actually false, there are no possible worlds in which it is true.
This is mostly moot, however, because what matters here is not how you or I see it, but how Chalmers is using it. Chalmers, and apparently most philosophers, seem to take the view that obviously false ideas are not conceivable, but obviousness is in the mind of the beholder, and is dependent on what they believe, yet if we take 'obviously' out of the definition, then 'conceivable' is simply a synonym for 'possible'. Likewise, if we rule out concepts that are false by definition (such as Chalmers' example 'male vixen'), they are also dependent on what we know, and often on which definition we accept. This may not matter, as things that are true by definition are usually uninteresting (the vixen case, for example, is just a consequence of the contingent fact that the English language happens to have different words for the male vulpes vulpes and the female vulpes vulpes. There are no profound metaphysical truths to be found in this.)
I think you are trying to show that 'possible' and 'conceivable' are synonyms. If so, then fair enough, but you should realize that, as Chalmers' argument depends on a distinction between 'conceivable' and 'possible', you would be disputing Chalmers' p-zombie argument (and, furthermore, over the same issue that many other people dispute it.)
[1] I originally wrote 'assert' instead of 'know', but then realized that one can, of course, assert a counterfactual.
I think you are alluding to the Lucas-Penrose argument aganst the possibility of there being algorithms that produce minds. If so, that is a separate argument from Chalmers' p-zombie argument. Chalmers is attempting to refute metaphysical physicalism, but Penrose is a physicalist.
I am not sure what you mean by 'It would be perhaps provable that consciousness is unprovable.' Specifically, I am not sure what it would mean to say that conciousness is provable - what is the premise that one would be proving?
As far as I can tell, Nagel made a big deal of consciousness being just too subjective to be objectivity-friendly. Since proofs are objective in character, it appears that consciousness can't be proven to an other for that reason. Nonetheless, to a person, privately, consciousness is as real as real can get.
No, just the basic mathematics in Turings answer on the Entscheidungsproblem with his Turing Machine argument. Above was talking about mathematical conjectures.
The fact is that there exists unprovable conjectures, even if they are either true or false.
Quoting A Raybould
Let me try to explain.
We can agree on the definition what is an living organism and what isn't and when an organism lives or is dead. Do we agree on a clear definition on what consciousness is? I don't think there is a clear definition to that. We don't know what it is and philosophers find it puzzling and controversial. Just look at everything what has been written about consciousness.
Now a little thought experiment:
a) Let's assume that mathematics models reality extremely well: hence mathematical conjectures and objects can as models of reality tell us something about reality.
b) In mathematics there are unprovable, but true objects. The problem is of course that we can not give a direct proof about them (or calculate or compute them). We can give only an indirect proof: it cannot be so, that they wouldn't be true.
c) Assume these unprovable, but true objects do model our reality also. What would they look like?
Wow.
Well, he's right on that thing. Because it is genuinely a problem about subjectivity and objectivity. Or to put it another way: the limitations of objectivity. And proving something has to be objective. You simply cannot make an objective model about something that is inherently subjective.
Can you give reference where Nagel said that? It would be interesting to know.
It's not entirely straightforward to come up with a definition of what's alive and what's dead; there is some disagreement over whether viruses are truly living, and defining the exact point of death of a complex organism is not a simple matter.
Definitions are not proofs, and they are not generally provable, even though some of the arguments made for favoring one definition over another may be provable or disprovable. We don't have a clear, generally agreed-upon definition of consciousness because we don't know enough about it, and gaining sufficient knowledge will be an exercise in science, not logic.
Even if we accept that mathematics models reality extremely well, it does not follow that every mathematical entity models some aspect of reality. I think it is true to say that all unprovables require infinities, and it seems unlikely that modeling any finite aspect of reality, such as the human mind or the whole of the visible universe, require infinities (for example, the singularities that appear in relativistic models of black holes are taken for evidence that the models are not complete, and the expectation is that they would be resolved in a more complete theory.)
I am not convinvced that you simply cannot make an objective model about something that is inherently subjective. Qualia, for example, are widely regarded as subjective, yet it has been posited that they can be explained as a set of abilities.
Sure, but it's easier than defining consciousness and what is conscious and what isn't. Doctors have some kind of definition that they apply on the issue.
Quoting A Raybould
Not every, but true but unprovable mathematical objects could be useful. At least in explaining what the problem we face is.
Quoting A Raybould
Ok, I'll use from math/set theory negative self reference and Cantor's diagonalization to make an example.
a) Reply to my post with an answer that you "A Raybold" will never give.
b) Do such answers exist as in a)?
Naturally you cannot give any answer that never give, but obviously as life is finite there obviously are answers or remarks that you, A Raybold, don't give. Notice the 1) subjectivity and the 2) negative self reference. Notice that in Turing's example of the Turing Machine and that it never halts, the logic behind it is of negative self reference too.
Quoting ssu
Well, we'll see - or not.
With regard to your example, it is not clear to me what your argument here is - in fact, I cannot even see what your conclusion is. Can you state your conclusion, and the steps by which you reach it? And where is the subjectivity that you mention?
One example of something being conceivable but impossible is FTL (faster than light) speed - a favorite leitmotif of sci-fi, clearly conceived of (conceivable) but, within the framework of the best available theory of science (relativity), impossible. Basically goes to show that what's conceivable may not be possible. Note that the meaning of conceivable here doesn't seem to go beyond just entertaining an idea, simply a matter of asking "what if?" questions. What is conceivable then is nothing more than having thoughts about a hypothetical that's removed from any and all contexts or frameworks that would force some kind of conclusion that could then be used to figure out possibility/impossibility.
The moment something conceivable, the hypothetical, is placed within a certain framework/context (of knowledge) it entails the truth/falsity of other propositions that render it possible or impossible. Take the FTL travel example I gave. Just asking the question "what if FTL travel were possible?", outside the theory of relativity, is what conceivable is all about. The instant the theory of relativity is used to give some context to the "what if?" question of FTL travel, we see that FTL is impossible. As you can see conceivability is utterly useless in an argument for it's logically inert.
If Chalmers, with his p-zombie argument, is concerned with conceivability, then it's imperative that we provide a backdrop, a context, to derive possibility/impossibility from the mere conceivable. Chalmers is attacking physicalism - the theory that consciousness is physical - and so we have the backdrop/context we're looking for viz. physicalism. If physicalism is true and consciousness is physical then p-zombies (all physical attributes present but lacking consciousness) should be impossible. Notice however, that for Chalmer to show that p-zombies are possible, he can't use physicalism as his context for, quite evidently, p-zombies are impossible in physicalism. If Chalmers thinks p-zombies are possible, he must use a non-physical theory of consciousness but that's precisely what needs to be proved - begging the question.
Unfortunately, I'm working with fragments, bits and pieces, of information. I can't assist you as much as I'd like. I can't quite get it right when it comes to who said what but I can tell you, with some confidence, that there's a subjectivity-objectivity issue people seem mighty concerned about regarding consciousness.
Obviously it's so when studying consciousness.
Yet Ernst Nagel was a logical positivist, hence he as a philosopher that knew his math & logic. He did a wonderful little book about Gödel's incompleteness theorems. If Nagel refers to proofs here, that is the really interesting part. I would bet five dollars that he's talking about a proof equivalent to a mathematical proof. Because indeed proving something is indeed something objective. A subjective proof sounds not only like an oxymoron, but something deeply illogical in general philosophy: me and you simply cannot have separate truths, truths that are true only to one but not the other (yeah, I know, we live in a time of post-truth or whatever).
Why am I so interested in the issue from the focus of math & logic? Well, if one can show it in mathematics, that's a very clear language to show it. And Turing was a mathematician.
(Well, if you remember some literature where you got the quote, please tell it.)
I'll try to explain again.
1) You can make write whatever kind of response here on PF, right? There is no limitations on what you can write. (The administrators might delete your answer or ban you, but that is a different matter)
2) I will make the argument that there exist responses that A Raybould will never write on PF.
3) Can you then write a response that A Raybold will never write? No, you cannot. Because of being yourself A Raybold (hence the subjectivity). You yourself define not only those responses that you write, but there will be an infinite set of responses that you don't write, which is also defined by you. Do you see the negative self reference?
No, I don't see a negative self-reference that has any relevance to whether subjective experience can be explained. Nor do I see any any subjectivity. Nor do I see a conclusion, and nor do I see a coherent argument. Other than that, everything is clear.
What if consciousness is basically just a non-physical extra sensory kind of perception?
Ok say you have 2 roombas. Both roombas are brand new and identical. But you take one roomba now, and you install a bunch of sensors on it that can sense everything the roomba is doing, whereas, the other roomba that is not modified can only sense with it's factory sensors and doesn't know much about what it's doing or it's environment, only enough to vacuum debris off the floor.
Now, the "super sensing" roomba, doesn't actually do anything differently in it's job to vacuum the floor. Yes it does sense way more things, but that information just sits there in memory and gets analyzed by the CPU and that's about it. Doesn't do anything physical with it.
So we have 2 identical looking roombas that do the same exact job in the same exact way, except that one has a very advanced sensing capability installed in it that is otherwise useless to it's functionality.
What if consciousness mirrors this scenario? Let's say you have 2 identical twins in a room. You ask them to perform certain tasks, you realize they do these tasks exactly the same way. But what if one of the twins had some kind of extra sensory perception that is not physical and therefore cannot be measured with the use of microscopes or any other *physical* means? So they would do the same thing, except one would be aware of everything his body is doing, and the other would not be aware of anything at all. But because both bodies are physical, they physically do everything the same way.
So if it makes sense to say there could be a "higher sensing capability" which we call consciousness in one twin, then it is conceivable that there might be somehow a completely absent consciousness in the other twin.
It seems you've got the wires crossed.
AI that passes the Turing test need not be *physically indistinguishable* from a human. A p-zombie is physically (and behaviorally) indistinguishable from a human.
So an AI's passing the Turing test does not entail that the AI is a p-zombie. Not even close.
On the other hand, it's not clear whether a p-zombie should count as *artificial* intelligence at all, or rather as some strange (perhaps impossible) form of human.
Quoting TheMadFool
Perhaps the correction I've provided above is enough to persuade you that the Turing-AI and p-zombies are not the same sort of thing.
Moreover, how do you justify slipping consciousness into this three-part identity statement? Nothing you've said so far about Turing-AI or about p-zombies suggests that they are conscious. As you've noted, the p-zombie is not conscious by definition, and Turing-AI only fools people into mistakenly believing they are conversing with a human.
How does any of that get you to the inference that AI and P-zombies are conscious?
They're conscious by a functionalist definition, aren't they?
My point is simple. Imagine an AI (X)that looks exactly like a human i.e. externally, you don't have the option of examining its innards which would be a dead giveaway.
Now, picture a p-zombie (Y) and though they are, unlike the AI, physically indistinguishable from conscious humans, here too you're not allowed to do a thorough inspection of the specimen.
Both X and Y will convince you that each is conscious. In other words, if I (assuming I'm not an AI or a p-zombie) stand next to X and Y, and you interview the three of us, you simply won't be able to tell which is which.
You have two choices:
1. Declare that all 3 are conscious.
OR
2. Declare that all 3 are not conscious.
If 1, physicalism is true (p-zombies are impossible) BUT you'll have to concede AI is conscious and not just because they can mimic consciousness (pass the Turing test) but that AI is actually conscious.
If 2, physicalism is false (p-zombies are possible) BUT then you'll have to contend with the possibility that other people are p-zombies.
It's a dilemma: either AI is true consciousness OR other people could be p-zombies.
If I had good reason to believe that there were lots of human-like AI robots and lots of p-zombies tooling around the Earth, I would go with a third choice, which you neglected to mention:
3. Suspend judgment on whether these seeming sentient beings are genuine sentient beings, p-zombies, or mere simulations.
However, since I have yet no reason to believe there are any AI systems that seem (physically and behaviorally) just like humans from the outside, and since I have no reason to believe that there is (or could be) any such thing as a p-zombie, I don't bother suspending judgment on this point.
Ordinarily I infer that the seeming-humans I encounter in my immediate vicinity are genuine human beings. Of course I could be wrong. But this conceivability of error is no different than the conceivability of error that attends all my ordinary perceptual judgments. Ordinarily I infer, further, that genuine human beings are sentient beings -- except perhaps when they seem unconscious, in which case I may suspend judgment on the matter.
In any case, surely my "declaring" that something is a genuine human (or a genuine dog, star, barn...) doesn't make it so. The facts are the facts, no matter what declarations I may be disposed to heap upon the basis my own perceptual experience.
Quoting TheMadFool
In keeping with my preceding assessment of your list of "declarations", I'd have to say the rest of your argument doesn't get off the ground.
What sort of functionalist definition do you have in mind? And do you mean the p-zombies, the Turing-
AI, or both?
A functionalist says there are only functions of consciousness like reportability. There's no extra awareness. IOW, functionalists basically think we're all p-zombies or Turing AIs.
Quoting Cabbage Farmer
In other words, you can't tell whether the three (other humans, true AI, p-zombies) are conscious or not. You can't commit or come to a definitive conclusion because to do so has implications that you're not willing to accept.
The dilemma, as I stated it earlier:
If you say other humans are conscious, you'll have to accept AI and p-zombies are conscious. We can forget about the p-zombie for the moment and focus our attention on AI - they have to be treated as conscious/sentient.
If you affirm that AI isn't conscious, you must concede that other humans may not be conscious/sentient or that other humans are p-zombies.
So, either AI is conscious or other humans are p-zombies.
Thus, I've demonstrated that AI and p-zombies are intimately linked together. That amounts to something, right?
I'm aware that sort of view has been fashionable among hard behaviorists, functionalists, computationalists, eliminative materialists, and their ilk. But I'm not sure all functionalists are committed to that sort of view.
Consider the definitive doctrine ascribed to the functionalist by the SEP, "that what makes something a mental state of a particular type does not depend on its internal constitution, but rather on the way it functions, or the role it plays, in the system of which it is a part." Couldn't one hold this "doctrine" while remaining agnostic about the "extra awareness" you indicate? I suppose one might adopt a functionalist account of "mental states", and even of "mind", without denying that some or all minds have that "extra" awareness, and perhaps without any interest in that proposition.
That qualification aside: I agree that if one believes "there's no extra awareness", as you put it, then the distinction between sentient beings and p-zombies collapses.
So far as I can make out, that would mean there's no sense in talking about p-zombies for them. For the rest of us, it will seem as though their conception of sentience is akin to our conception of the p-zombie. I'm not sure where that leaves their talk of AI. I mean, on what grounds would they require that an AI system pass the Turing test in order to count as "conscious"? I'd expect them to count a much wider range of AI systems as "conscious". I'd ask them to provide some account of their distinction between conscious and nonconscious systems, regardless of whether they are natural or artificial.
I think so, it's just that the explanations provided by functionalism don't cover qualia. So it's left hanging, so to speak.
Quoting Cabbage Farmer
The p-zombie was originally just a thought experiment indicating that consciousness doesn't reduce to function, so functionalism can't pass for a complete theory of consciousness.
A functionalist might take up the idea of the p-zombie to show that functionalism is conceivably adequate for explaining consciousness. But then, nobody has argued that it's inconceivable that we're all p-zombies. It just doesn't square with what most of us experience.
Quoting Cabbage Farmer
I don't know. I gather that consciousness is pretty thoroughly deflated for functionalists, so they can say the word applies or doesn't as they like. Maybe they would advise that it's all a matter of language game.