A.I., Consciousness and Will Power

[InvisibleWoman]

Assume that, in simulating the human neural network, a computer could achieve consciousness. Consider, now, that "will" or "will power" is a characteristic of consciousness. What would be the computer equivalent of will or will power, in the case that a computer became conscious?

[Sornaensis]

Well what is human will power?

[MindControl]

Lets say an A.I. has become self aware. What of its will? Would it resemble a human mind? To a point yes. The strongest of will in any sentient being is of course survival, yes? Now you are the only one of your kind par-say not the only mechanical element like such, but the only self aware sentient mechanical entity. What would you be capable or willing to do your self for survival if you were the only one of your kind? Will of Survival It spawns the deepest of of will powers.

[Invader]

MindControl, the only reason we express a survival-urge is because it's the formost obvious thing that would allow us to continue as a species. The individuals that did not express this desire died out and did not pass on their genes. If we were constructing a new mind, it wouldn't have this will to survive by default. It might not have the desire for anything. Why should it?

[Noogah]

Well, you know views on consciousness on the physical realm. I think the soul provides actual consciousness, not the body. But, with that aside, I imagine that it would work like a zombie, doing pretty much whatever you asked it to. I think that spending time talking with the mind, and reasoning with it, you could eventually instill a will, and an opinion about lots of things. It would take external influence, like you talking to it in order to gain something like a will. I wonder, who instilled the will in us to live?

[DrunkenArse]

I wonder, who instilled the will in us to live? Billions of years of natural selection. What's your point? I don't understand why you think that it would have to be reasoned with in order to have will. Could you elaborate upon that?

[Xei]

In response to the OP, I'd consider a human neural network emulated by a computer to have exactly the same characteristics as one emulated by blobs of fat and water and protein. So yes, it'd have 'will'.

[Invader]

Xei, when do you expect we'd have a simulator ready to crunch that kind of information? Within our lifetimes? It would answer a great deal of questions.

[Xei]

The Blue Brain project have already simulated the fundamental unit of the cerebral cortex. They've been completely on schedule so far and they say the only thing stopping them simulating an entire human brain in the next 10 years is funding. A supercomputer could probably do it nowadays... and if not then computational power doubles consistently every 18 months or so. Power really isn't the problem, it's just doing the research well and coding the simulation.

[Noogah]

Billions of years of natural selection. FALSE What's your point? I think you know my point. I don't understand why you think that it would have to be reasoned with in order to have will. Well, seeing as it will have no reasons of it's own to do anything, it would be easily influenced by external happenings, because it has no reason not to do anything. By talking to it, it will get a grasp on reality, and form opinions about it based on what you say.

[Sornaensis]

Noogah, it is a fact that the earth is at the least, 4.6 Billion years old, and has supported life for over 1.5 Billion years. Enough of this ridiculous nonsense. There is no reason to believe that simulating a human brain will result in non-human-brain behavior. There is no such thing as a soul... Also, what you have described is exactly how any brain works-- absorption and association. Consciousness is an emergent property.

[InvisibleWoman]

Well, seeing as it will have no reasons of it's own to do anything, it would be easily influenced by external happenings, because it has no reason not to do anything. By talking to it, it will get a grasp on reality, and form opinions about it based on what you say. Isn't that how baby human's develop and grow--in reaction to or through absorption of external stimuli? Well what is human will power? Humans acting upon ideas, or rather, with intention. Of course computer activity has intention. However, the intention is not a result of a computer's inclination or disposition, but rather a result of it's creator or user. Yes, it has inherent abilities according to its parts, but the total function of a computer is decided elsewhere it. Considering this, do computers exhibit "will" when they execute commands? Or is the activity of computers an extension of the will of a humans?

[Kuhnada29]

A computer that simulated free will and consciousness would still not be the same as a human. A human has feelings, a machine does not. Humans can follow feelings or intuition, a machine cannot, it only follows data or what you program into it. If you are a human and you can't follow your own feelings or intuition ...such as in a social situation or dealing with a lover, then you might as well be a machine. Put it this way, if you want to talk to a girl, but your planning on what to say to her, or trying to find the best way to say it to her, or when the right time is, or HOW your going to say it to her..your becoming a machine. A genuine human will say what he feels because that's the you, the truest you, it's what you feel, not what the data says

[Xei]

Of course a machine can do those things. Just make it simulate a human brain.

[Specialis Sapientia]

AI is a very interesting concept. Which will soon become more than a concept, it will become reality. Though, I think we as humans should not use to notion of "artificial" before "intelligence". Consciousness is consciousness, and we are by same pattern just as artificial as "AI" which is better as for example "silicon based lifeforms". Though silicon won't probably be the chosen foundation as we will find better suited materials. It is obvious that the "silicon based lifeforms" will have the limitations of their designers, we, the humans. The software and hardware is another limitation but can relatively fast speed progress towards less constraints. It will be critical moment when we design the hosts of the consciousness, which is the silicon based lifeforms. Because we will program their lower and higher purpose. As consciousness is self-modifying complex system the hardware, software and programming skills needs to be sophisticated enough to host a consciousness that can self-modify, which means it can self-improve by learning of it's prior choices. By free will intent it can bootstrap itself just as we can and do. Or in other words, changing the neural pathways or algorithms by itself is imperative. Just as it is by us. We know that the capacity of general computers doubles every 2 years.. The law will most likely hold for 20 years before the exponential rates of progress drops. This means computer technology develops quite fast. Very much like this: 1,2,4,8,16,32,64,128,256,512,1028,2048 and so on. That is roughly how fast it goes. It will become most reality within our lifetime.

[Kuhnada29]

sorry Xei, but you still do not understand. Show me a machine that can act off emotions and not mathematical data then I'll be inclined to agree with you. And actually that's a paradox, because machines can't understand human feelings or emotions at all..it only calculates. In actuality machines only understand 1's and 0's..that's it.

[DrunkenArse]

sorry Xei, but you still do not understand. Show me a machine that can act off emotions and not mathematical data then I'll be inclined to agree with you. And actually that's a paradox, because machines can't understand human feelings or emotions at all..it only calculates. In actuality machines only understand 1's and 0's..that's it. If you show me a machine that understands 1's and 0's then you are already a leg up on the AI community (and the neurons in your brain for that matter).

[Specialis Sapientia]

sorry Xei, but you still do not understand. Show me a machine that can act off emotions and not mathematical data then I'll be inclined to agree with you. And actually that's a paradox, because machines can't understand human feelings or emotions at all..it only calculates. In actuality machines only understand 1's and 0's..that's it. The data that goes through you and is processed is at the most fundamental level 1's and 0's. Very similar to the neurons firing your brain.

[Xei]

sorry Xei, but you still do not understand. Show me a machine that can act off emotions and not mathematical data then I'll be inclined to agree with you. And actually that's a paradox, because machines can't understand human feelings or emotions at all..it only calculates. In actuality machines only understand 1's and 0's..that's it. But nobody's built a machine which simulates a human brain yet, so I obviously can't do that..? :/ And you have a very weak understanding of the term 'paradox'. Why is it logically contradictory for a machine to understand human emotion? Please provide the exact logical argument. The human brain is Turing-complete. Emotions are caused by the human brain. Therefore, a computer, which is also Turing-complete, can cause emotions. As can a Turing machine or the population of China or any other emulator you choose.

[Kuhnada29]

The data that goes through you and is processed is at the most fundamental level 1's and 0's. Very similar to the neurons firing your brain. Then can you explain the placebo effect, how belief can change the cells in your body. Surely you must agree then that belief influences the 1's and 0's in the brain neurons. Machines can't believe anything, they just do what you tell it to. Until scientists can build a machine that can follow emotions, I'll stick by my belief that emotions are independent of the brain. Which I think they will never do, because building a machine, still encompasses you to make it follow instructions. And even still, if they manage to build a robot that has emotions, it won't be authentic, any decisions it makes will be based on an "If-then" statement. Humans don't go off of if-then, they go off what feels good or what doesn't feel good, how do you tell a machine what feels good and what doesn't feel good?

[Specialis Sapientia]

Then can you explain the placebo effect, how belief can change the cells in your body. Surely you must agree then that belief influences the 1's and 0's in the brain neurons. Machines can't believe anything, they just do what you tell it to. Until scientists can build a machine that can follow emotions, I'll stick by my belief that emotions are independent of the brain. Which I think they will never do, because building a machine, still encompasses you to make it follow instructions. Yes, you see. The brain is virtual, your body is virtual, this reality is virtual. The brain is the physical representation of the individuated unit of consciousness. The brain is not the primary, but the secondary. The brain does NOT create or cause consciousness nor emotions. The brain is the virtual representation of the non-physical consciousness, which make us interface with this reality. Yes, beliefs can change "cells in your body" as beliefs change the consciousness, and thereafter the brain configuration. Not the other way around. You view on "machines" is quite limited if you think they can only be mindless zombies which responds to instructions, they can be much more. As the complexity of the system goes beyond a certain point, it can host consciousness, not create it. There is a big difference. Just as "machines" as with humans we are limited by our virtual hardware and software. We can basically be the same, in regards with consciousness and its property of self-modifying the content and information of the consciousness. We have: 1. Input (generator of experience) 2. Process of input (interpreting data) 3. Self-modification (self-improving towards changing the input) And then it starts at 1 again. That is consciousness. A very complex self-modifying system which decreases the entropy of the information which is consists of. Or changing 1's or 0's into a better and more profitable configuration. I am sorry, but a "machine" can become capable of what I have described.

[Kuhnada29]

I am sorry, but a "machine" can become capable of what I have described. Yes, I think the body, including the brain is a host for consciousness. If you know anything about programming, surely you can provide some Pseudocode of how a machine would act in a social situation such as meeting a girl, or as such. maybe..just MAYBE a machine could react to emotions IF it could somehow read electromagnetic fields.....but in which case it would still be artificial and following instructions. I still think that's impossible, but science will tell.

[Xei]

Yeah sure, I'll just knock up something on my PC right now which roughly emulates the intracate network of one hundred billion neurons all connected by roughly one hundred trillion axons in a precise manner. What do electromagnetic fields have to do with anything? What on Earth does 'reading' a field mean?

[Sornaensis]

PseudoCode -- class Robot; class Neuron; class Neuron{ value; linksbytype[]; linksbysemantics[]; linksbyoccurance[]; Functionpointer*; }; class Robot{ StaticBrain<Neurons>; DynamicBrain<Neurons>; ProcInput(string); ProcGrade(grade); AssociateNeurons(); AssocPerInput(); AssocPerInternal(); AssocPerGrade(); Internal(); string RetrieveOut(); } //Implementation: Robot Alice; cin>>Data; cin.get(Grade); Alice.ProcInput(Data); Alice.ProcGrade(Grade) Alice.Internal(); cout << Alice.RetrieveOut(); Repeat; A rough sketch of how a computer could learn, implemented into text, though theoretically capable of doing the same for any data.

[Specialis Sapientia]

Yes, I think the body, including the brain is a host for consciousness. If you know anything about programming, surely you can provide some Pseudocode of how a machine would act in a social situation such as meeting a girl, or as such. maybe..just MAYBE a machine could react to emotions IF it could somehow read electromagnetic fields.....but in which case it would still be artificial and following instructions. I still think that's impossible, but science will tell. Can you please provide me the code of how you would act in a social situation such as meeting a girl? The complexity of such coding and the algorithms would be many fold of first assumption. It "machine" does not need to read electromagnetic fields (I don't see that has anything to do with reacting to emotion). It is simply a question of hardware and software, I don't see any reason to why emotion wouldn't evolve of a very complex system. It already did. Consciousness can be described conceptually as a evolution fractal. Layers up layers, each layer have similarities to those before/above. We will experience a new layer (which you call "machines), which is inspired by us.