Practically perfect............This follows a few of my own theories on AC. Consciousness will never arise in digital or any serial computer...ever! And as Jeff points out, the mere fact that computers are of Von Neumann architecture negates any possibilities of anything we would qualify as intelligence let alone consciousness. Look forward to further developments on his site.
From what I recall, instead of using transistors, they're supposed to rely on quantum effects to establish their ones and zeros. If so, they're still essentially digital, and I would assume they'd utilize the Von Neumann architecture. If so, I don't think this would make any difference regarding their ability to be conscious - at least in the same way that the human brain achieves it.
If a machine can persuade me that it is abstracting, rather than just calculating, then I will be happy to ascribe consciousness to it. Almost certainly some sort of recursion is involved, some process which creates a model of itself and is able to associate that model with its own identity. This is a tall order, and still may not get us to the desired result, but I think it's a good goalpost.
@LordSlag From the computer's perspective, if you could somehow instantaneously stop the program and examine the values of all of the transistors corresponding to the values in the registers, it would be nothing more than a lot of digital ones and zeros. There would be nothing amidst all of this that would be much different if it was simulating, say, an oil refinery or a weather pattern.
@LordSlag The way I see it (and I may be wrong), a digital simulation would in essence be equivalent to a gigantic spreadsheet of values; values pertaining to quantities like voltages, concentrations, velocities, spatial coordinates, time, etc. I have a hard time understanding how all of these "values in memory registers", albeit changing very rapidly, can be equivalent to something like an actual sensation of the smell of peppermint. From the computer's perspective... (see next comment).
neuroscience is likely our best guide, perhaps coupled with psychology and philosophy of mind (including the fact that humans do not have complete autonomy - freewill. only when we ignore reality, and even those imaginations are grounded in prior experiences)....all put into the language of mathematics.
I think qualia are only part of the problem and perhaps a relatively small part. Many species respond to stimuli in complex ways, do sophisticated pattern matching, dream, and display behavior that would seem to indicate that they feel emotions. Whether they are conscious is another question. In my opinion consciousness has much more to do with the abstract. What humans do which most other species do not is to "step outside" the world of stimuli and access the abstract universe of ideas.
neural simulation is close enough. I think just a matter of time until GPGPU are able to simulate around 100million neurons, we might be able to see self learning machines in the future
Machines can never become aware as they can only function simultaneously with the passage of present time. There is no space or time between the past and the future in which awareness can exist. Awareness requires the ability to straddle the micro-minuscule moment of "now" with minute extensions into both the past and future.
Perspective if irrelevant. From my perspective, if I freeze your brain and then take a close look, there's no qualia there either. But, if I put you back together flawlessly and revive you then you regain your consciousness. Even if you looked at a functioning brain or simulation, the biological/electromagnetic functioning doesn't reveal the content of consciousness, nor would a simulation of such a brain that is sufficiently detailed to replicate those functions. Cogito ergo sum, sir.
no. but without qualia (in where the self functions as an object to be conscious of), you would achieve a pure consciousness that isn't moved by anything. Without senses, it has nothing to move it. compare to freewill debates. we interact with an environment, creating a sort of give and take. we have a reason for doing something, on behalf of something (experience, memories of qualia). without this, you have a human in sensory deprivation. nothing moves it, no WHY i do this. conscious OF nothing
You fail to take into account physics governed simulations of biological reality. Simulating a human brain down to the atom/photon scale would result in consciousness regardless of what sort of computing platform it was done on. Consciousness is not predicated on qualia, that is an unfounded assumption.
further, without something to consider true (believe), you have no HOW to do something on behalf of why (why i do something and then how i do it). there is no reason to choose a over b if a and b do not exist. people have to be taught how to think, to manipulate memories of experiences. it's why we have school, it's our jobs as parents. as long as we are modelling computerized robots after human beings and our understanding of consciousness, it will require us to first understand ourselves.
hey maybe there is no such thing as a consciousness or self identity in objective reality. maybe machines are already conscious in a way we cant understand. i think that any agent that can detect its surroundings and act accordingly is conscious. some machine have been able to do that
Practically perfect............This follows a few of my own theories on AC. Consciousness will never arise in digital or any serial computer...ever! And as Jeff points out, the mere fact that computers are of Von Neumann architecture negates any possibilities of anything we would qualify as intelligence let alone consciousness. Look forward to further developments on his site.
From what I recall, instead of using transistors, they're supposed to rely on quantum effects to establish their ones and zeros. If so, they're still essentially digital, and I would assume they'd utilize the Von Neumann architecture. If so, I don't think this would make any difference regarding their ability to be conscious - at least in the same way that the human brain achieves it.
If a machine can persuade me that it is abstracting, rather than just calculating, then I will be happy to ascribe consciousness to it. Almost certainly some sort of recursion is involved, some process which creates a model of itself and is able to associate that model with its own identity. This is a tall order, and still may not get us to the desired result, but I think it's a good goalpost.
@LordSlag
From the computer's perspective, if you could somehow instantaneously stop the program and examine the values of all of the transistors corresponding to the values in the registers, it would be nothing more than a lot of digital ones and zeros. There would be nothing amidst all of this that would be much different if it was simulating, say, an oil refinery or a weather pattern.
@LordSlag
The way I see it (and I may be wrong), a digital simulation would in essence be equivalent to a gigantic spreadsheet of values; values pertaining to quantities like voltages, concentrations, velocities, spatial coordinates, time, etc.
I have a hard time understanding how all of these "values in memory registers", albeit changing very rapidly, can be equivalent to something like an actual sensation of the smell of peppermint. From the computer's perspective... (see next comment).
neuroscience is likely our best guide, perhaps coupled with psychology and philosophy of mind (including the fact that humans do not have complete autonomy - freewill. only when we ignore reality, and even those imaginations are grounded in prior experiences)....all put into the language of mathematics.
I think qualia are only part of the problem and perhaps a relatively small part. Many species respond to stimuli in complex ways, do sophisticated pattern matching, dream, and display behavior that would seem to indicate that they feel emotions. Whether they are conscious is another question. In my opinion consciousness has much more to do with the abstract. What humans do which most other species do not is to "step outside" the world of stimuli and access the abstract universe of ideas.
What are your thoughts on quantum computers?
neural simulation is close enough. I think just a matter of time until GPGPU are able to simulate around 100million neurons, we might be able to see self learning machines in the future
Machines can never become aware as they can only function simultaneously with the passage of present time. There is no space or time between the past and the future in which awareness can exist. Awareness requires the ability to straddle the micro-minuscule moment of "now" with minute extensions into both the past and future.
@jeffkosmo
What do you think about quantum computers?
I haven't gotten you wrong, you're making stuff up.
Perspective if irrelevant. From my perspective, if I freeze your brain and then take a close look, there's no qualia there either. But, if I put you back together flawlessly and revive you then you regain your consciousness. Even if you looked at a functioning brain or simulation, the biological/electromagnetic functioning doesn't reveal the content of consciousness, nor would a simulation of such a brain that is sufficiently detailed to replicate those functions. Cogito ergo sum, sir.
no. but without qualia (in where the self functions as an object to be conscious of), you would achieve a pure consciousness that isn't moved by anything. Without senses, it has nothing to move it. compare to freewill debates. we interact with an environment, creating a sort of give and take. we have a reason for doing something, on behalf of something (experience, memories of qualia). without this, you have a human in sensory deprivation. nothing moves it, no WHY i do this. conscious OF nothing
You fail to take into account physics governed simulations of biological reality. Simulating a human brain down to the atom/photon scale would result in consciousness regardless of what sort of computing platform it was done on. Consciousness is not predicated on qualia, that is an unfounded assumption.
further, without something to consider true (believe), you have no HOW to do something on behalf of why (why i do something and then how i do it). there is no reason to choose a over b if a and b do not exist. people have to be taught how to think, to manipulate memories of experiences. it's why we have school, it's our jobs as parents. as long as we are modelling computerized robots after human beings and our understanding of consciousness, it will require us to first understand ourselves.
hey maybe there is no such thing as a consciousness or self identity in objective reality.
maybe machines are already conscious in a way we cant understand.
i think that any agent that can detect its surroundings and act accordingly is conscious.
some machine have been able to do that