Why computers will never be conscious
- Thread starter Fen
- Start date
- Status
AndersHermansson
Registered Senior Member
Sounds to me like a feeble attempt to explain something in scientific terms. There's nothing in scientific theory that prohibits machines from being conscious. In fact, for a machine to be a conscious in a human fashion all you need is to emulate a body and an hook it up with reality. If we will be capable to do such a thing in practice? Who knows.
I have a point for the room.
If the positions and connections of all the neurones in a human brain were recorded and entered into a neural net emulator (or a computer hardware net, if possible) then surely the emulator would behave in exactly the same way as the person whose brain was donated?
You would of course have to ensure that all the correct inputs were linked up.
I did not know that true randomness existed? I acknowledge 'Brownium motion' (spling?) for example as apperaing random, but if you had all the kinetic data I think you would be able to predict the motion.
Similarly, if you knew all the neuronic connections (and inputs) inside a brain, could you not predict the thoughts and subsiquent changes in a persons net?
We are just interacting matter after all?
If the positions and connections of all the neurones in a human brain were recorded and entered into a neural net emulator (or a computer hardware net, if possible) then surely the emulator would behave in exactly the same way as the person whose brain was donated?
You would of course have to ensure that all the correct inputs were linked up.
I did not know that true randomness existed? I acknowledge 'Brownium motion' (spling?) for example as apperaing random, but if you had all the kinetic data I think you would be able to predict the motion.
Similarly, if you knew all the neuronic connections (and inputs) inside a brain, could you not predict the thoughts and subsiquent changes in a persons net?
We are just interacting matter after all?
Siddhartha
Registered Senior Member
The Heisenberg uncertainty principle injects a degree of true randomness to everything.
If there were no randomness at all, that would mean that there is no free will, as every action you make is the result of neurons firing in reaction to stimuli which was set in motion 15 billion years ago. You are simply physics then... which is all possible. But if there IS randomness...things change dramatically.
-AntonK
I didn't know that. Is it absolutely certain that this principle shows true randomness? or just true uncertainty?
That is correct, no free will as such. At least in my opinion. I have been told however, that you cannot take the data and work backwards, but that is obviously not too relevant.
What about the first half of my post?
could someone further explain how true randomness allows free will? the way i see it, determinism is still valid whether an individuals actions are determined by predictable patterns, or unpredicable randomness. determinism in a general sense implies predictability, so perhaps randomness invalidates determinism... but i dont see free will as becoming true by default.
Well, Good question about the neural net. My belief is that we are missing something in our Neural Net simulators. In most neural nets I've seen, there has been a static architecture where all neurons are set. The weights between connections can change, and this is the "learning" this works well for pattern recognition. BUT...the human brain also changes its connections, and its initial architecture is FAR more complex even from birth.
My question to people that are trying to come up with human level intelligence from neural nets is, if neural nets are so great, why are humans the ONLY neural net (brain) that has intelligence at our level. What's the defining factor that makes us so much more intelligent than any other animal?
-AntonK
kmguru
Staff member
That is because, humans have language processing system where as other animals have none or rudimentary processes. The built-in Bayesian type network is integrated to the language and thought (awareness) with integrated connections. A neural network's output depends on the type of inputs. Most of the today's NN is limited to one or two datatypes where as humans are exposed to a lot more both internal and external.
That is a small part of the big picture.
That is a small part of the big picture.
There is no "true" randomness. A word about Heisenberg's Uncertainty Principle, you are correct in assuming that it talks about uncertainty not randomness, but aren't they one in the same? Randomness is something that's outcome is not certain. And certainty has no randomness. Now, I'm not trying to argue that they are the same, but in this instance I think we can say they are fairly similar.
Now, the only reason randomness exists in the first place isbecause of Heisenberg's Principle. If it was untrue, and we could measure both the position and the momentum of particles with complete accuracy, everything could be predicted. So Blue: yes you are correct and incorrect at the same time. There is no free will, but, since Heisenberg's Principle is in place, our inability to predict the path of particles (cool alliteration) makes it impossible for us to determine it. So basically, randomness is our inability to decipher the fate of particles because of Heisenberg's Uncertainty Principle.
'Nonquantized' storage, means no one has measured the total amount of storage, so it is unknown. It has been misapplied here.
Language is a form of data compression, which can be effective if used with reasonable care.
Cogito cogito ergo cogito sum. (Ambrose Bierce)
I think that I think, therefore I think that I am. Another fun way to look at the world.
'Hey don't blame me! I'm the monkey, not the organ donor!' (who cares?)
Language is a form of data compression, which can be effective if used with reasonable care.
Cogito cogito ergo cogito sum. (Ambrose Bierce)
I think that I think, therefore I think that I am. Another fun way to look at the world.
'Hey don't blame me! I'm the monkey, not the organ donor!' (who cares?)
Last edited:
If a computer system is self-aware (it IMHO can be), isn't it also conscious?
I would say a system could evolve consciousness based on an understanding that the physical world that we all know follows a destined outcome through a mixture of causality and the lack of control over parody.
People will still debate that they aren't bound by fate, but thats because they "Aren't conscious" of their decision making following such a pattern.
I state this just to point out that you can say a machine won't gain consciousness, but what about yourself? Are you really as conscious of the universe as you might think?
People will still debate that they aren't bound by fate, but thats because they "Aren't conscious" of their decision making following such a pattern.
I state this just to point out that you can say a machine won't gain consciousness, but what about yourself? Are you really as conscious of the universe as you might think?
kmguru
Staff member
When people say the machine will never gain consciousness, what they are really saying is that the machine can never be a human. It is like saying an apple can never be an orange. Logically they are right. That is the end of the argument, unless they define the word differently like apple is a fruit, can orange be a fruit?
Actually kind of adding to that statement Kmguru, I've always wonder the point of the Turing test/ Loebner prize.
My concern is quite simple, As human beings we don't like to be lied to, So why are people trying to develop "Intelligence systems" that "lie" about being human and try to mimic humans to the extent of not being distinguishable from us.
This in fact could be a potential danger, imagine a computerised system that is merged with some form of monitoring system like "Carnivore", it could be possible to completely fool someone that the "Intelligence System" is someone else.
Great for social engineering, but bad for business.
My concern is quite simple, As human beings we don't like to be lied to, So why are people trying to develop "Intelligence systems" that "lie" about being human and try to mimic humans to the extent of not being distinguishable from us.
This in fact could be a potential danger, imagine a computerised system that is merged with some form of monitoring system like "Carnivore", it could be possible to completely fool someone that the "Intelligence System" is someone else.
Great for social engineering, but bad for business.
The Evil Sponge
Registered Member
ok, I've heard a lot about this, take this into consideration:
Assuming there is a higher intelligence unreachable that we cannot fully comprehend (what we would call god), we were given the miracle of life, at first it was very basic (one celled organism) and then became more advanced through BILLIONS of years. The pre homosapien at some point must have snapped out of it and laughed or seen something never felt before, the ability to make a desicion and view things from outside that box of instinct.
Now a computer has been around since early machines, very basic like a calculator, within this short of time, it has become what we rely on to do a lot of our basic every day necessities. WIth the aid of us programming AI and all of these other high tech and complex programs, we are acting somewhat like a God. S. to say that a computer can never acheive consciousness is faulty in my opinion because with the intelligence that they have, one day one might take a look outside of what it is programmed to do and have a "laugh".
Now give a computer race a billion year to evolve. ..
Assuming there is a higher intelligence unreachable that we cannot fully comprehend (what we would call god), we were given the miracle of life, at first it was very basic (one celled organism) and then became more advanced through BILLIONS of years. The pre homosapien at some point must have snapped out of it and laughed or seen something never felt before, the ability to make a desicion and view things from outside that box of instinct.
Now a computer has been around since early machines, very basic like a calculator, within this short of time, it has become what we rely on to do a lot of our basic every day necessities. WIth the aid of us programming AI and all of these other high tech and complex programs, we are acting somewhat like a God. S. to say that a computer can never acheive consciousness is faulty in my opinion because with the intelligence that they have, one day one might take a look outside of what it is programmed to do and have a "laugh".
Now give a computer race a billion year to evolve. ..
I agree with The Evil Sponge. Computers are exponentially increasing in computing power every year or so. Even if computers never become sentient, we will continue to build faster computers and more flexible programs. Eventually, computers will be able to handle environmental interaction with at least the processing power of a single human. At this point, even if computers are only programmed to respond a specific way to a specific stimuli, they will become indestinguishable from man (to our perception). We may not have progressed humanoid robotics to the point where we can simulate the human body, but the female Customer Service agent you speak to on the phone my be nothing more than a box in a "clean room".
We may never be able to build computers that have human emotions, but does that make us superior? Should we really be trying to duplicate our primal emotions? Would we possibly be better off without them? Well, maybe not, I would miss curiosity.
I hated the end of bicentennial man. I did not understand why he would try to become more human. If he was lonely, fine, build more units similar to himself. He obviously had the knowhow to do the engineering (not to mention there would be schematics somewhere). Do not get me wrong, I do not consider man inferior to machines or vice versa, I just consider the two different species each with their own strengths and weaknesses.
Personally, I do not think an infinite storage capacity or processing ablility is needed for conciousness to develop. It only needs to be enough to handle the workload. I do not understand why it would not be possible (even preferable) for machines to download their daily memories/ thoughts into a central master system that crunches the data and feeds it back in a managable size (for instance, the computer downloads all of the recorded digital video data, but receives it back in a downsized version (without as much detail). After said amount of time the central system quits returning files that are never accessed.
On another note: Uncertainty and Randomness are not the same thing. Randomness means an equal chance that any particular value will be true. Uncertainty deals with the unknowable.
- KitNyx
We may never be able to build computers that have human emotions, but does that make us superior? Should we really be trying to duplicate our primal emotions? Would we possibly be better off without them? Well, maybe not, I would miss curiosity.
I hated the end of bicentennial man. I did not understand why he would try to become more human. If he was lonely, fine, build more units similar to himself. He obviously had the knowhow to do the engineering (not to mention there would be schematics somewhere). Do not get me wrong, I do not consider man inferior to machines or vice versa, I just consider the two different species each with their own strengths and weaknesses.
Personally, I do not think an infinite storage capacity or processing ablility is needed for conciousness to develop. It only needs to be enough to handle the workload. I do not understand why it would not be possible (even preferable) for machines to download their daily memories/ thoughts into a central master system that crunches the data and feeds it back in a managable size (for instance, the computer downloads all of the recorded digital video data, but receives it back in a downsized version (without as much detail). After said amount of time the central system quits returning files that are never accessed.
On another note: Uncertainty and Randomness are not the same thing. Randomness means an equal chance that any particular value will be true. Uncertainty deals with the unknowable.
- KitNyx
- Status