difference between brains and computers

[gib65]

I had an idea. I want to know what people think of it:

 

Do you think that one of the main reasons computer engineers have been having trouble mimicking the human brain is simply because computers are digital while brains are analogue?

[Callipygous]

i think its because evoltion is WAY cooler than intel.

 

 

but uh... analogue? no idea. maybe : P

 

i thought it had more to do with something like one of them being a multitasking system and the other only doing one at a time so quickly it seems like multitasking.

[invert_nexus]

That's definitely one of the problems, but not the only one. And not even the most limiting.

 

A huge problem with 'mimicking' the human brain is that no one understands how the human brain does what it does. We're on the verge of a huge leap in understanding, but at the moment (and for years to come... maybe) we are infants in our knowledge.

 

However, instead of trying to mimick the brain perfectly, they can instead focus on end results and allow the middle layers to take care of themselves.

 

One of the largest problems, I tend to believe, is that which is illustrated by Goedel's Incompleteness Theorem.

 

Computers are the quintessence of formal systems and as such are bound to get caught in endless loops of logic. Humans (and other thinking creatures) are informal systems which can exit the loop when they wish.

 

No blue screens of death for humans.

[Callipygous]

suicide. psychosis.

 

but yeah, we are definately much less limited.

[invert_nexus]

Aye. We do suffer from our problems, but they are not of the same type. Not infinite loops.

 

An infinite loop would not be suicide. It would be nothing. It would be nothing.

 

You could say that there are instances of psychosis (catatonia, hysterical paralysis, Parkinson's disease as in the movie Awakenings, and others) that could be seen as a sort of infinite loop. But it's not really. It's brain damage. Not a problem with the system itself.

 

This is a very difficult concept to talk about. And I'm no math expert which is where it's illustrated most fully so I lose a lot in my interpretation of the explanations. But if you would care to know more, I can recommend a site or two.

 

First, Godel, Escher, Bach is a most interesting book that delves into the topic.

 

Also, John Lucas wrote an excellent series of papers on the subject you might find some interest in. Here's a site where both his articles and those rebutting him are gathered together. And here's his home page.

 

It's really some slippery stuff to get your mind around, but very interesting.

 

Also, Turing was working along similar lines as Godel when he come up with his Halting Problem.

 

I highly recommend GEB (Godel, Escher, Bach). It's a long read but one of the most intelligent and imaginative books I've ever read. Truly worth the read.

[bascule]

The problem is this:

 

Consciousness is complicated, and hard to understand

 

That is all. The material implementation is inconsequential; we merely need to figure out the mathematical model.

[Tommahawk]

No, I think it is because the mathematical representation of the brain and how it functions is not known and therefore cannot be re-created by computers.

[Nevermore]

A brain is much like a quantom computer, considering many things at once. So once we figure out how to build a quamtom computer, (which is expected to be in the next few decades.)...

[dna mauro]

the difference is that we have emotions,that is impossible for PC right now.In brain we can make new ideas and toughts without needing any sourc code or orders.it is theoretically possible to make a artifectical brain,using carbon,need to put atoms together to make neurological system,where is power in,but in that brain cant learn or forget also it doesent have emotions,that control brain a little.also its hard to conect it somehow with pc or something to use it.only if we can make half biological brain(that is able to create new connections and brake these in brain),it would be super,but it still doesen have emotions,that are controlled by hormons,that react only with certain molekuls or places in brain.hope you understand something.

[uncool]

DNA, I believe that, should we be able to find all the reactions in the brain/body, it would be a simple (though extremely tedious) task to replicate a human. All chemical reactions can be modeled by switches of a type or computers, I believe, and therefore, I believe that eventually, computers will be able to create "emotions" of a type. Our brain has its own source code that tells it to create new code - and that has been done with computers, just not quite to the same extent. Sorry for rambling here!

-Uncool-

[invert_nexus]

Uncool,

 

So. You're a determinist then?

Did you read those links to Lucas's arguments against determinism, by any chance?

 

And what of chaos theory? And quantum theory?

The brain is nothing like a computer. A computer is digital. The brain is analog. I'm a believer that a brain couldn't be modeled efficiently by any system less complex than itself. Brain-like behavior could be implemented (think Chinese Room and/or black box) but it would not be needful to model the brain precisely. The brain is a work of neurons and neurotransmitters. Action potentials and ion channels. The brain is the result of its medium and it works because of it. To try to implement its function precisely in another medium (electronics) would entail certain translations of form that might easily invalidate that structure.

 

You should utilize your medium for its benefits and limitations. Flesh and circuitry are not the same and it would be foolish to treat them as such.

 

 

Dna Mauro,

 

By a similar token, emotions are an end product of various processes. They are a motivational factor for more complex behaviors. They can be simulated by various means. Think of the Hitchhiker's Guide to the Galaxy and their "Genuine People Personalities." The first developed was boredom. Once they switched from using a herring sandwich (if memory serves) the rest followed in quick succession. Colin the euphoric robot comes to mind.

 

The point where emotion takes on a new tinge... a more human tinge... is in the interpretational process. And it is here that that the elusive factor of 'consciousness' (human consciousness... I'm not going to get into the problems of nomenclature right now) lies.

 

Heh. Lies. That's what the interpretive mechanism is adept at. Lying. Telling stories. Explaining. Rationalizing.

 

When you can make a machine that lies in order to explain its functioning in the absence of knowledge of that functioning, then you'll be close to 'it'. When you can make a computer that seeks meaning and isomorphisms in the products of its senses, then you'll be on the road to awareness. Self-awareness. When you've designed a machine that can revel in contradictions and explain them away with a wink and a nod, then you'll be approaching human. And not until then can we say that the difference between brain and computer is being bridged.

[uncool]

invert nexus, that is exactly what I meant with the Chinese room. I was just saying an extremely inefficient way in which human brain behavior could be modeled - but much less efficient than most ways.

Now, to read the paper against determinism...thanks for the link.

-Uncool-

[dna mauro]

i understand what you want to say,problem is that, i know too less english words to write my toughts in here.

[Cris]

The Brain is analog? Are you sure? Each neuron takes a number of inputs from many other neurons and then may or may not fire a pulse to another neuron - that sounds very digital to me.

 

Note that each neuron fires on average 300 times per second. And with 200 billion of them operating in parralel that gives an equivalnet computing power of around 20,000 x 3GHz Intel Pentium processors.

 

The overwhlming reason we haven't come close to mimmicking the brain is that we have yet to put together any type of processing system that powerful. It'll take about another 10-15 years.

 

And there is nothing special about emotions - they are still functions of the brain.

[invert_nexus]

The Brain is analog? Are you sure? Each neuron takes a number of inputs from many other neurons and then may or may not fire a pulse to another neuron - that sounds very digital to me.

 

Hmm.

You have a point there. Neurons do fire once the firing threshold has been reached. On or off. With no real variations in the action potential (although the firing threshold does change as the neuron 'tires out'.)

 

But, it is also analog in that the communications between neurons is through neurotransmitters and not action potential. The chemical messengers between cells is analog. And throw in glial cells regulating the neurotransmitter level in any given synapse even more (a new idea which is not wholly proven yet, but is certainly tantalizing in its evidence) and you have a whole spectrum of values rather than on or off.

 

So. As has been said, the brain is complicated. It's sort of a hybrid creature. Digital and analog.

Interesting.

[mmalluck]

Lets not forget about parallel processing.

 

Computer are very good at processing code line by line in a linear fasion. We can use time slices (I.E. Computer, I want you to run this code for application A for X amount of time and then run the code for apllication B for Y amount of time) giving the appearance of parallel processing.

 

You brain on the other hand can do dozen of things at a given time (taking care of vital functions, walk, talk, and chew gum at the same time) without having to switch between these functions.

 

Hurray for specialized areas of the brain!

[dna mauro]

http://www.med.harvard.edu/AANLIB/cases/caseNA/pb9.htm look at this link, it shows, the main functions of brain(or something,not so good in english)

 

didnt find any links, where was about making brain of atoms,thats possible(it is possible to put carpon atoms together,so how we want them to be, it also transfers energy,that is one important thing for brain)

[pnaermao]

It has been mentioned that the brain works both as a digital and analog machine. Also, it was mentioned of the numerous variables and possible options that can be done, not just "on and off" like a computer. But there is another aspect that must be considered because of the brain's organic nature. When a specific area of the brain is used continually, the neurons themselves begin to shift around and form new connections to make these often used areas easier and easier to use. So, can it really be compared to a machine? Until we are capable of building a computer that has the ability to re-wire its own hardware to produce the most effective system available, could we ever come close to creating A.I. or self-awarness? Or does this contribute to the brain's abillity to create its own "code". Maybe this is the source of a living organisms ability to be intuitive, creative, and "out of the loop". Think, for every thought there is the possibility of a completely unique system behind it.