Is it possible for AI to be conscious?

@asdf2000: Because it seems like the easier job to do.

@Rapoza: Yep, that is so. Also, don't forget qualia (Q) is not detachable from cognition (say M). I can think and at the same time experience qualia. This constructs consciousness (C), or C = M ~ Q.

--- Nuked ---

But it doesn't seem to require consciousness to do anything. If anything, the ego that is associated with human consciousness tends to get in the way of accomplishing the things that we want deep down.

--- Nuked ---

--- Nuked ---

Just because an AI is able to accelerate our evolution or solve problems in a clever way doesn't mean the AI has a mind or is conscious. AI use formal symbol systems. The AI or CPU or whatever is syntactical and does not use semantics to arrive at an output.
This may seem off topic but it will get at the root of what I am talking about.

What a word refers to is not the same as it's meaning. This is most obvious example is with proper names but is true of more abstract things.....Barrack Obama, The President of the United States, and the tall dark skinned man wearing a suit. All of these pick out and refer to the same person but have different meaning. So its easy to see that Barrack Obama picks out the specific person as it is his proper name, but when I say "POTUS" did X today you understand who I am talking about because you understand the meaning when I say POTUS. Consider if you didn't understand the meaning of any words at all and then looked in the dictionary to understand what Y word refers to, under definition Y=ABCD. So to understand Y you then have look up ABC and D, which will lead you in a infinite loop. So why does this matter?

When a computer or and AI does something it uses a formal symbol system to get an output. Symbol systems take a symbol ( that has no meaning for example X) and manipulate the symbol with other symbols and syntactical rules to produce a new symbol output(x^2+y^2=z^2). The rules for symbol systems are based on the shape of the symbol and not the meaning of the symbol. Now I am just starting my cognitive science degree but from my current understand of neural networks they are using a complex math function to arrive at an output. Even for classification problems like is this a tree or is it a rock the class is translated into a numerical value. So while this method is valuable and will like improve life it doesn't mean that the computer performing this is conscious or that the computer has a mind. The AI may be able to correctly identify what the words POTUS refers to but it does not understand the meaning of the words since its only capable of syntactical manipulation. So while it is then possible to create an AI that is able to simulate our behavior we shouldn't be confused into thinking that this simulation is a duplication since the process at which it arrives at the behavior is so much different then ours.

So would you (or whomever) say that we could not replicate the "process" that is consciousness (any level of complexety, assuming other organisms have conciousness) performed by what we would call a robot ?


Would you also say that humans have a ceiling as to what they can replicate synthetically ? Given that we have been synthetically replicating our own biological system step by step over time, what we call "machines" a way, or more vaguely technology as a whole.
And at that point, would we say that conciousness is apart of our biological process, or are consciousness and biology apart?

I think the question is not "if" , but "how".

IMO

I think these answers are extremely narrowsighted, and in reality are only answering "could my Laptop be conscoius?"

Why would you assume that in a distant future AI can only work through input A - Response B and in no way could remotely understand semantics, I mean, we have evidence that RNA evolved into conscious beings with a very simple survival of the fittest process and yet you cant imagine a way a process like that could ever happen in any other form or organism that isnt protein based? come on.

Very nice explanation from FMLuser about the current way of working of AI, thanks.

Baalim, nobody is narrowsighted, its just our understanding of consciousness is too limited as of 2016. I can imagine a future world, in which we are all only energy. The process of how to reach that world is what we are lacking.

On April 20 2016 00:35 Rapoza wrote: I can make an analogy that 2 different persons cannot have the exact same brain state, yet, they might share the exact same consciousness about something thou their Qualia will be different.

Yeah, I knew you were going in that direction, and that is why I told you that you can't separate M and Q.

J. Kim believes that we can have scenario like the one you suggest:

Are mental properties physically reducible? Yes and no: intentional/cognitive properties are reducible, but qualitative properties of consciousness, or 'qualia,' are not.

The problem is that he was strongly criticized for that, because such variant will separate the unity of consciousness. Here is the explanation in short:

1) qualia (Q) is multiple realizable and epiphenomenal
2) the rest (M) are reducible and genuine properties

Let's consider the following case:
P1 = M1 [~ Q] -> P2 = M2

If we want M1 to be a subset of Q we have a paradox:
P1 and Q are different properties (from 1 and 2), P1 is identical with M1 and both share the same domain (from 2), so M1 can't be a subset of Q. Thus at time t we have multiple mental states.

The other case is the current M state to be interrupted every time a Q state appears or vice versa:
P1 [~ Q] -> P2 = M2 -> P3 [~ Q] -> P4 = M4 -> ...

So if we want to have epiphenomenal qualia in reductionist view, we are faced with two incoherent options - multiple mental states or some kind of interruption.

On April 20 2016 08:42 Expiate wrote: Baalim, nobody is narrowsighted, its just our understanding of consciousness is too limited as of 2016

Then it doesnt make any sense to say it cannot happen

On April 20 2016 00:48 Rapoza wrote: Show nested quote + On April 20 2016 00:22 asdf2000 wrote: But it doesn't seem to require consciousness to do anything. If anything, the ego that is associated with human consciousness tends to get in the way of accomplishing the things that we want deep down. The consciousness, alongside creativity and many other things, is why we want things in the first place, otherwise we would be as active as a rock. For example, we need creativity solutions to our current society problems. The AI as we know would just answer: "Recycle your stuff". If an AI could think, its possible to reach a reasonable and accurate conclusion much much faster because it can process and simulate data better then our limited chemical brain.

I'm not sure consciousness is why we want things, but rather a manifestation of our wanting of things.

I see what you are saying, but I am still not convinced consciousness has a role to play in AI.

Actually, I feel like it kind of takes the problems with the complexity involved in programming a strong AI and tries to bypass them by going "eh, well just figure out how to make it conscious". I don't think that's relevant to the strengths of AI. I believe that strong AI will be about writing an extremely complex elegant program that is recursive. It will be able to evolve it's own code at an exponential rate. Whether or not it's conscious will be irrelevant (and probably unknowable).

On April 21 2016 00:41 asdf2000 wrote: Show nested quote + On April 20 2016 00:48 Rapoza wrote:   On April 20 2016 00:22 asdf2000 wrote: But it doesn't seem to require consciousness to do anything. If anything, the ego that is associated with human consciousness tends to get in the way of accomplishing the things that we want deep down. The consciousness, alongside creativity and many other things, is why we want things in the first place, otherwise we would be as active as a rock. For example, we need creativity solutions to our current society problems. The AI as we know would just answer: "Recycle your stuff". If an AI could think, its possible to reach a reasonable and accurate conclusion much much faster because it can process and simulate data better then our limited chemical brain. I'm not sure consciousness is why we want things, but rather a manifestation of our wanting of things. I see what you are saying, but I am still not convinced consciousness has a role to play in AI. Actually, I feel like it kind of takes the problems with the complexity involved in programming a strong AI and tries to bypass them by going "eh, well just figure out how to make it conscious". I don't think that's relevant to the strengths of AI. I believe that strong AI will be about writing an extremely complex elegant program that is recursive. It will be able to evolve it's own code at an exponential rate. Whether or not it's conscious will be irrelevant (and probably unknowable).

Well if the compuer becomes an all-knowing demi-god wouldnt itself perfectly know if it is conscious or not under our own narrow definition of the word?

I think these answers are extremely narrowsighted, and in reality are only answering "could my Laptop be conscoius?" Why would you assume that in a distant future AI can only work through input A - Response B and in no way could remotely understand semantics, I mean, we have evidence that RNA evolved into conscious beings with a very simple survival of the fittest process and yet you cant imagine a way a process like that could ever happen in any other form or organism that isnt protein based? come on

well sure, but it's important to keep in mind that we have made exactly 0 progress towards this over last 100 years of computer science. When people talk about AI and its potential they often assume that we are discussing something that is at least remotely relevant to the body of knowledge we call AI research, and also to the algorithms this research program has generated over the years. For strong AI we need something fundamentally different from the computers we have. While it is reasonable to speculate that in the future someone will come up with a technology literally no one can imagine the details of today, it is a distant dream, something like colonizing Alpha Centauri or whatever.

On April 19 2016 19:50 Baalim wrote: Show nested quote + On April 19 2016 13:04 Stroggoz wrote: Human's are conscious in every way that machines can't be. That's the one question here that is simple to answer Yet you didnt answer it, so please let us know, in what ways humans are conscious that machines can never be?

machines are just objects which take inputs and outputs. There is no reason to think they can be conscious like a human being, especially when science hardly understands how human beings are conscious.

I agree with a mode of reasoning that the more unlikely a hypothesis is, the more reason you need to believe it.

On April 20 2016 23:04 Baalim wrote: Show nested quote + On April 20 2016 08:42 Expiate wrote: Baalim, nobody is narrowsighted, its just our understanding of consciousness is too limited as of 2016 Then it doesnt make any sense to say it cannot happen

yeah it does because there's no evidence to show that it can happen, and it's simply ridiculous to suggest we could create strong AI with such a limited understanding.

On April 21 2016 23:24 Stroggoz wrote: Show nested quote + On April 19 2016 19:50 Baalim wrote:   On April 19 2016 13:04 Stroggoz wrote: Human's are conscious in every way that machines can't be. That's the one question here that is simple to answer Yet you didnt answer it, so please let us know, in what ways humans are conscious that machines can never be? machines are just objects which take inputs and outputs. There is no reason to think they can be conscious like a human being, especially when science hardly understands how human beings are conscious. I agree with a mode of reasoning that the more unlikely a hypothesis is, the more reason you need to believe it.   On April 20 2016 23:04 Baalim wrote: Show nested quote + On April 20 2016 08:42 Expiate wrote: Baalim, nobody is narrowsighted, its just our understanding of consciousness is too limited as of 2016 Then it doesnt make any sense to say it cannot happen yeah it does because there's no evidence to show that it can happen, and it's simply ridiculous to suggest we could create strong AI with such a limited understanding.

No it doesnt unless we were talking about a specific time frame like, "in the next 100 years".

Does a computer have consciousness? Obviously not ...

But just wait til the singularity. There's a good chance it happens in your lifetime if you are under 40.

--- Nuked ---

bump
http://www.iflscience.com/brain/man-missing-most-of-his-brain-challenges-everything-we-thought-we-knew-about-consciousness/all/
interesting article

If the AI starts beating FlaSh's T as Zerg then I would agree that it can be conscious