Poll: Is artificial awarenesss possible?

[Deleric]

An artificial intelligence involves thinking. To create a fully working humanoid thought process, along with memory, self awareness, creativity and problem solving, you'd also need shitloads of other addons that can stimulate the system in such a way that a real though process can be formed. Off the top of my head, you'd need:

Prominence: Something that could tamper with memory to create certain connotations in certain events or ideas. This could also create notions of importance, irrelevancy and attachment, which would then in turn create feelings of stress, relaxation and hate/love.

...Or maybe I'm just pulling this shit out of my ass.

 

[Maze1125]

I believed I made myself clear but let me elaborate. There is a very distinct difference from Artificial Intelligence and Artificial Awareness. If the definition is correct, I personally believe we can make robots do independent thinking. But Awareness is the ability to have an abstract thought about your existence. I don't know where it is located but somewhere we know that we are human. You have a conscious, the smartest computer in the world does not. How can you possibly program that? Copying neurons like above posters have said, would not work. Since the neuron that moves your arm does not make you conscious. For the gamers that play Mass Effect out there, All VI like Avina are artificially intelligent. All Geth and EDI are artificially intelligent and also have Artificial Awareness.

Your argument is basically "I have no idea how someone could give a computer awareness, therefore it can't be done."

 

[Shurikens and Lightning]

I believed I made myself clear but let me elaborate. There is a very distinct difference from Artificial Intelligence and Artificial Awareness. If the definition is correct, I personally believe we can make robots do independent thinking. But Awareness is the ability to have an abstract thought about your existence. I don't know where it is located but somewhere we know that we are human. You have a conscious, the smartest computer in the world does not. How can you possibly program that? Copying neurons like above posters have said, would not work. Since the neuron that moves your arm does not make you conscious. For the gamers that play Mass Effect out there, All VI like Avina are artificially intelligent. All Geth and EDI are artificially intelligent and also have Artificial Awareness.

Your argument is basically "I have no idea how someone could give a computer awareness, therefore it can't be done."

Your argument is basically "I also can't prove something, but this guy has to be wrong because it is the Internet!"

Who knows, maybe in a thousand years we will get somewhere with Artificial Awareness. But in the next 20, 50, 100 years? Most likely not. How can we teach a robot something that we cannot even understand, let alone program. Besides "I think. Therefor I am" we have a little less than basic knowledge about Self Awareness. Example: " How do you prove a robot knows it's a robot, and be correct?" It's near impossible since I can't even prove you are self aware, because you are all robots! I could make a computer answer yes when somebody asks if it is Self Aware. It still is not self aware and is nothing more than a bucket of bolts and wires.

 

[Eclectic Dreck]

I believed I made myself clear but let me elaborate. There is a very distinct difference from Artificial Intelligence and Artificial Awareness. If the definition is correct, I personally believe we can make robots do independent thinking. But Awareness is the ability to have an abstract thought about your existence. I don't know where it is located but somewhere we know that we are human. You have a conscious, the smartest computer in the world does not. How can you possibly program that? Copying neurons like above posters have said, would not work. Since the neuron that moves your arm does not make you conscious. For the gamers that play Mass Effect out there, All VI like Avina are artificially intelligent. All Geth and EDI are artificially intelligent and also have Artificial Awareness.

My assertions are quiet simple - first, while it is entirely possible to model the basic operation of a neuron mathematically and this idea has been leveraged in a wide variety of applications to simulate a mind more advanced than a cockroach is beyond the technical capibilites of any available computer systems. Second, human history is full of instances where someone declared a problem to be beyond human capacity to solve which were later proven incorrect. Every single advance in scince in 200 years is a perfect example. Third, this claim is being made out of a lack of adequate information - specifically it is unknown at this time precisely where "awareness" comes from. Unlike fields of mathematics and physics, sciences where the major questions remaining lie in the resolution of quantum mechanics and gravity and a wide variety of limited applications of theoritical mathematics, biology and psychology present common and persistant gaps in knowledge.

As such, while you assertion that artifical awareness is impossible is seemingly reasonable, history indicates that the problem can be solved so long as there is sufficient will to do so. Moreover, the tendancy to claim a problem cannot be solved is born from ignorance and in this fashion you join the ranks of intellects so prestigeous as Newton who presumed that a mechanical model of the solar system must rely on regular divine intervention in order to remain stable when, in fact, he simply needed a different way to approach the problem.

 

[Woodsey]

I don't see why not.

We can do AI, and we can do (quite limited) interaction with machines - that Honda robot thing, anyone?

It can detect and remember objects, and has basic interaction. As someone already said, its less about the materials (computer chips vs. the crap we've got floating around) then you'd think.

 

[Maze1125]

Your argument is basically "I also can't prove something, but this guy has to be wrong because it is the Internet!"

No it isn't.

For a start, that's a Burden of Proof fallacy. My argument doesn't need proof, but yours does.
I'm not arguing that it definitely can be done, just that you have failed to give any good reason that it can't be beyond "Well I have no idea how to do it."

Who knows, maybe in a thousand years we will get somewhere with Artificial Awareness. But in the next 20, 50, 100 years? Most likely not.

Now you're completely changing your argument.

How can we teach a robot something that we cannot even understand, let alone program.

We can't. But that doesn't mean we will never understand it, and once we understand it, it shouldn't be long before we could understand how to program it.

Besides "I think. Therefor I am" we have a little less than basic knowledge about Self Awareness. Example: " How do you prove a robot knows it's a robot, and be correct?" It's near impossible since I can't even prove you are self aware, because you are all robots! I could make a computer answer yes when somebody asks if it is Self Aware. It still is not self aware and is nothing more than a bucket of bolts and wires.

Indeed, that is the conundrum of conciousness.
But, even if that were never solved, that still doesn't mean we couldn't give a computer awareness. We would never be able to prove we had, but we still might have done it.

 

[Threesan]

Assuming awareness arose via evolution, it should be fundamentally possible to create a new awareness. Evolution is not magic, it just had a lot of time and space on its hands.

Even a quantum computer---if it is equivalent to a finite non-quantum computer at least---will not be capable of awareness,

Quantum computers are most definitely not equivalent to non-quantum computers.

Even a non-deterministic Turing machine can be run on a deterministic Turing machine. Nothing's changed on what is possible, only on what is feasible (by changing some complexity classes). Thus, equivalent.

Rather, I would reject the notion that human awareness is capable of perfect infinite self-reference, and that a program cannot access any instruction it wants or make any observation on its behavior it cared to. Amongst other things in the wall/link.

 

[Nivag the Owl]

I think it's totally possible. Robots can easily be programmed to "learn" using trial and error programming. If you program a robot to solve a a rubix cube. Not by giving it the moves and sequences, but programming it to be able to function the rubix cube, and then giving it the purpose of solving it. It can totally learn how to.

Memory is very easy and hardly needs an explanation in robotics. Self realization can be imitated but whether or not it can look at it's own parts and recognise them as a part of itself is difficult.

Creativity is the one I'm thinking is most unlikely.

 

[Trotgar]

It's probably possible, but only after we're much more advanced.

I think it should somewhat represent the functionality of a brain, and that would be very complicated.

 

[Katherine Kerensky]

I'll just give the simple answer that I give to these kinds of situations:
Nothing is impossible, just extremely improbable.
So yes, I believe it is possible. Don't know what it would take to make it a reality though.

 

[Admiral Stukov]

I belive that our awareness and well, most of our mind is "software" and thus wheter or not the hardware is organic or synthetic would not matter.
I do not belive in the Terminator style robot uprising tho, I mean you should keep better control over your self-aware computer if it controls a bunch of nukes.
So well it's just a matter of time.
Now I'll go sit in a corner and wait for someone to come up with a way to transfer my mind to an android body.

 

[Maze1125]

Even a quantum computer---if it is equivalent to a finite non-quantum computer at least---will not be capable of awareness,

Quantum computers are most definitely not equivalent to non-quantum computers.

Even a non-deterministic Turing machine can be run on a deterministic Turing machine. Nothing's changed on what is possible, only on what is feasible (by changing some complexity classes). Thus, equivalent.

Each bit in a computer has two possible outcomes.
Each q-bit in a quantum computer has an infinite number of possible outcomes.

Therefore they are not, in any way, equivalent.

 

[TheRightToArmBears]

I think it's possible, but I can't see why it would be a good idea.

 

[Shurikens and Lightning]

snip

Your asking me to prove the future? That's a pretty hard task to do. A debate about whether something in the future will happen is on par with "Is there a god?" We can spend all day arguing or we can just realize that these are interpretations.

This debate is very simple but you are determined to ruffle some feathers, do you think Artificial Awareness is possible? I say no, you say yes. Then you followed up by calling my opinion wrong? I don't know why you are dragging out this, but you are determined to fight about opinions and not hard facts.

 

[Maze1125]

snip

Your asking me to prove the future? That's a pretty hard task to do. A debate about whether something in the future will happen is on par with "Is there a god?" We can spend all day arguing or we can just realize that these are interpretations.

This debate is very simple but you are determined to ruffle some feathers, do you think Artificial Awareness is possible? I say no, you say yes. Then you followed up by calling my opinion wrong? I don't know why you are dragging out this, but you are determined to fight about opinions and not hard facts.

Don't pull the "it's just my opinion" card. That's bollocks.
You've been out right saying it's impossible and arguing for that claim, that not "just an opinion".

I say no, you say yes.

No, as I just explained, you say "no", I say "You have no good reason to claim that."

 

[Stoic raptor]

Possibly, but i doubt it will happen

 

[zehydra]

Even if it is possible to create self aware computers, you would never be able to prove that they were, in fact, self aware.

In fact, you cannot know for certain that anyone around you is self aware either, since you cannot know what another person is thinking.

 

[Velvo]

You know, the end of that wall of text said something like 'What hubris to think we can simply manufacture, or evolve, awareness with a pile of electrified silicon hardware and some software rules.' A more interesting point is, what hubris it is to believe that for some reason we humans are special enough to be the only kind of intelligence that could have an amount of awareness. All we are is some biochemical electrical programming making connections from inputs and running them around through millions of bioelectrical processors. WE evolved this way. The more and more we understand about what makes us aware, why couldn't we speed the process for another sort of intelligence?

For that matter, why couldn't we create a hybrid bioelectical and electrical? Use the sheer number of a neural net of bioelectrical processors and the great speed of electrical processors and have the best of both worlds? Lets make an android like Data!

 

[Threesan]

Even a quantum computer---if it is equivalent to a finite non-quantum computer at least---will not be capable of awareness,

Quantum computers are most definitely not equivalent to non-quantum computers.

Even a non-deterministic Turing machine can be run on a deterministic Turing machine. Nothing's changed on what is possible, only on what is feasible (by changing some complexity classes). Thus, equivalent.

Each bit in a computer has two possible outcomes.
Each q-bit in a quantum computer has an infinite number of possible outcomes.

Therefore they are not, in any way, equivalent.

By definition a qubit is a quantum representation of a bit, or binary integer. Measuring a qubit has exactly two possible outcomes: 0 or 1. Anything shifting the alignment away from exactly the angles representing 0 or 1 is error or randomness, and is discarded by the measurement. An n qubit quantum computer will then have 2[sup]n[/sup] outcomes, not infinite. However, I suspect you may have been thinking not of outcomes, but that somewhere in the process of computation, the (supposed) continuity of some quantum property could be exploited to do something special (not just faster, but, dare I say, magical).

This is no longer a point concerning quantum computing, but the distinction between discrete and analog computing. To exploit the (supposed) continuity of some physical property, we would need infinitely fine measurements, infinitely fine operators, and perfect protection from the rest of the universe. Lacking infinite precision, we can sufficiently (indistinguishably?) approximate analog computation with discrete computation. It may be slower, but it can be done.

Here is my (not impossibly mistaken) intuitive understanding of the operation of quantum computing:
Qubits can be initialized relative to measurement so that there is an equal probability of measuring a 0 or 1. All qubits initialized in this way then collectively have an equal probability of taking on any one of their collective 2[sup]n[/sup] possible values. Subsequent application of an operation will apply to all states in the superposition and eliminate from the initially 2[sup]n[/sup] possible states those states inconsistent with the application of the operation (this is accomplished using entanglement/destructive interference -- the operations by design establish entanglement links without collapsing the superposition). The final measurement will give (ignoring noise) one of the 2[sup]n[/sup] states that is consistent with the operations applied, assuming there is such a state.

That ability to operate on all 2[sup]n[/sup] states simultaneously is the power of QC. This can make it faster for certain tasks, but it doesn't address what is fundamentally possible.

TL;DR & PS

If human awareness had a knob that controlled speed of thought/perception of time, if you turn it up to 110%, you still have awareness. If you turn it down to 90%, or 50% you still have awareness. If you turn it down by a factor of a trillion (or some other obscenely more obscene number) -- still, awareness? Maybe not one that is practical or even meaningful in our existence, but it still is what it is.