• You are currently viewing our forum as a guest, which gives you limited access to view most discussions and access our other features. By joining our free community, you will have access to additional post topics, communicate privately with other members (PM), view blogs, respond to polls, upload content, and access many other special features. Registration is fast, simple and absolutely free, so please join our community today! Just click here to register. You should turn your Ad Blocker off for this site or certain features may not work properly. If you have any problems with the registration process or your account login, please contact us by clicking here.

Distributed Ne?

runvardh

にゃん
Joined
Jun 23, 2007
Messages
8,541
MBTI Type
INFP
Enneagram
6w7
Instinctual Variant
sx/so
How online games are solving uncomputable problems

Are these problems "uncomputable"? Or are we just currently not able to model them effectively?

Other applications which might lend themselves to this approach?

Computers still aren't able to pick the most likely adjustment with out trying them all. The word "uncomputable" is a mysnomer as the information can be calculated - it just takes a while. We humans, on the other hand, just by looking at something can "see" more likely results with out having to think through each possibility one after another. The issue is which is still more efficient and so the effectiveness.
 

Salomé

meh
Joined
Sep 25, 2008
Messages
10,527
MBTI Type
INTP
Enneagram
5w4
Instinctual Variant
sx/sp
Computers still aren't able to pick the most likely adjustment with out trying them all. The word "uncomputable" is a mysnomer as the information can be calculated - it just takes a while. We humans, on the other hand, just by looking at something can "see" more likely results with out having to think through each possibility one after another. So in short it's just effectiveness.

That's the point. How to model that effective edge. What these gamers are doing isn't strictly 'calculation', it's more complicated than that.

Computers are only restricted by the algorithms they are given to work with. I'm wondering if we might be able to be more intelligent than use "brute force" methods...what that might look like....
 

runvardh

にゃん
Joined
Jun 23, 2007
Messages
8,541
MBTI Type
INFP
Enneagram
6w7
Instinctual Variant
sx/so
That's the point. How to model that effective edge. What these gamers are doing isn't strictly 'calculation', it's more complicated than that.

Computers are only restricted by the algorithms they are given to work with. I'm wondering if we might be able to be more intelligent than use "brute force" methods...what that might look like....

Well, most of how our mental processing works starts with a database that gets filled from before birth. Mother nature took millions of years to get it right, God did it in 7 days. We've only been at it for 80 years.
 

Salomé

meh
Joined
Sep 25, 2008
Messages
10,527
MBTI Type
INTP
Enneagram
5w4
Instinctual Variant
sx/sp
It's filed as electrical impulses in the connections of our brain, something else we're still far from knowing much about.

I'm more a proponent of the tabula rasa, myself.
 

runvardh

にゃん
Joined
Jun 23, 2007
Messages
8,541
MBTI Type
INFP
Enneagram
6w7
Instinctual Variant
sx/so
I'm more a proponent of the tabula rasa, myself.

I never said we start off with it prior to neuron 1, but once the network is there information starts flowing and that gradually compliles. As it is the only synthetic analogue we have are some robots a guy put together that learn how to walk and navigate. This aquiring and storage of data requires a half decent amount of parallel processing and storage space. We're just not there yet with the hardware, never mind the fact that it still takes months of loading experience data to move around in the world. I have yet to see someone with the money who would really want to fund such a thing.
 

Simplexity

New member
Joined
Jul 15, 2008
Messages
1,741
MBTI Type
INTP
Hmm... are you meaning to attack this from a more cognitive science/psychology perspective. Do you think there is some form of algorithmic processing going on in our brain in addition to this blank slate. Or do you think our knowledge and intelligence is processed in a different manner.

I noticed runvardh that you mentioned it's stored in the form of electrical impulses, so do you think its more repetitious? in the sense that we keep adding deeper, more elaborate, and expansive layers and they get recorded in each instance forming the intricate web of knowledge that is our brain?

Are you trying to figure out if our mental processing can be deduced, or are you arguing it is much more complex and therefore can not be understood in mathematical terms?
 

ptgatsby

Well-known member
Joined
Apr 24, 2007
Messages
4,476
MBTI Type
ISTP
I never said we start off with it prior to neuron 1, but once the network is there information starts flowing and that gradually compliles.

Yup... the mind has built in pattern searching and processing systems in place automatically. Plus the raw power of our mind is many orders of magnitude stronger than the computers available.

The computational problems come from not having enough power and not powerful enough algorithms.
 

Simplexity

New member
Joined
Jul 15, 2008
Messages
1,741
MBTI Type
INTP
Yup... the mind has built in pattern searching and processing systems in place automatically. Plus the raw power of our mind is many orders of magnitude stronger than the computers available.

The computational problems come from not having enough power and not powerful enough algorithms.

So is intelligence a factor of differentiation in "pattern searching and processing ability?"

Can those things be learned to a certain extent?
 

runvardh

にゃん
Joined
Jun 23, 2007
Messages
8,541
MBTI Type
INFP
Enneagram
6w7
Instinctual Variant
sx/so
Hmm... are you meaning to attack this from a more cognitive science/psychology perspective. Do you think there is some form of algorithmic processing going on in our brain in addition to this blank slate. Or do you think our knowledge and intelligence is processed in a different manner.

I noticed runvardh that you mentioned it's stored in the form of electrical impulses, so do you think its more repetitious? in the sense that we keep adding deeper, more elaborate, and expansive layers and they get recorded in each instance forming the intricate web of knowledge that is our brain?

Are you trying to figure out if our mental processing can be deduced, or are you arguing it is much more complex and therefore can not be understood in mathematical terms?

It's basically the adding more layers into the web that makes this sort of processing possible. I haven't really put forth trying to figure it out, nor am I saying it cannot be understood. I am, however, saying that it's a bit more complex than we're able to work with right now.
 

Salomé

meh
Joined
Sep 25, 2008
Messages
10,527
MBTI Type
INTP
Enneagram
5w4
Instinctual Variant
sx/sp
I never said we start off with it prior to neuron 1, but once the network is there information starts flowing and that gradually compliles. As it is the only synthetic analogue we have are some robots a guy put together that learn how to walk and navigate. This aquiring and storage of data requires a half decent amount of parallel processing and storage space. We're just not there yet with the hardware, never mind the fact that it still takes months of loading experience data to move around in the world. I have yet to see someone with the money who would really want to fund such a thing.
This is just a thought experiment. Binary processing is the major constraint.

Hmm... are you meaning to attack this from a more cognitive science/psychology perspective. Do you think there is some form of algorithmic processing going on in our brain in addition to this blank slate. Or do you think our knowledge and intelligence is processed in a different manner.

I noticed runvardh that you mentioned it's stored in the form of electrical impulses, so do you think its more repetitious? in the sense that we keep adding deeper, more elaborate, and expansive layers and they get recorded in each instance forming the intricate web of knowledge that is our brain?

Are you trying to figure out if our mental processing can be deduced, or are you arguing it is much more complex and therefore can not be understood in mathematical terms?

Electrical impulses are that - impulses. They don't store data permanently. They handle communication. Neurons build connections and somehow data gets encoded biochemically. But that's a digression. I'm talking about processes, not data.

I'm not making an argument, I'm asking a question, well, several.

I guess I'm thinking about neural networks, AI, etc. I don't know much about that field.
 

ptgatsby

Well-known member
Joined
Apr 24, 2007
Messages
4,476
MBTI Type
ISTP
So is intelligence a factor of differentiation in "pattern searching and processing ability?"

Can those things be learned to a certain extent?

That's a tricky question. Intelligence is defined commonly as "general IQ", which leads to the problem of how effective it is at measuring holsitic intelligence. Normally fluid intelligence is made up of a battery of questions that test various things... but in reality, they are simply correlated factors. So the relationship isn't well defined in a way that can be broken down into neurological terms.

However, pattern searching is a very large component of intelligence yes, and likely the biggest part as a whole. There are other things though, like memory, abstraction... (crystalized intelligence, both part of general intelligence, is more present here) but many of those relate to variations on pattern searching (pulling the right memory, seeing how it applies...).

Can it be learnt... If we talk about raw power and makeup, it's not very flexible. But this is pattern searching we are talking about - from a neural net. It needs to be adaptable. So yes, it can be "learnt", but only in the sense of adaptation. The raw power, the... "cpu" part of the mind... that doesn't change significantly. It tends to create a ceiling on what the "adaptive" "software" does.
 

runvardh

にゃん
Joined
Jun 23, 2007
Messages
8,541
MBTI Type
INFP
Enneagram
6w7
Instinctual Variant
sx/so
I know only the barebones about AI and cog sci which is what I've been spouting only half successfully. They have figured out some simple stuff, but we're a long way off.
 

Salomé

meh
Joined
Sep 25, 2008
Messages
10,527
MBTI Type
INTP
Enneagram
5w4
Instinctual Variant
sx/sp
So is intelligence a factor of differentiation in "pattern searching and processing ability?"

Can those things be learned to a certain extent?

Intelligence takes many forms. Everything can be learned...to some extent.
 

Simplexity

New member
Joined
Jul 15, 2008
Messages
1,741
MBTI Type
INTP
No doubt. I think it's one of the biggest issues in the subject of cognitive science. I remember having a discussion about this a while back ago with somebody else on this forum who's majoring in that. It's pretty fascinating, if you try and attack it from the mathematical level and not the biological level. There two very different approaches to attacking it.

Do you think one would prove more fruitful to ultimately coming up with an answer. Obviously in the end we would have to attack it from the algebraic end if we are going to possibly try and integrate this with computers, but maybe the the hidden key lies in a different arena and that is why it is such a difficult topic to crack.

Heres a link to the thread that touches on some of this stuff if you're interested in some backgroudn reading. The computational stuff is near the end of the thread.

http://www.typologycentral.com/forums/philosophy-spirituality/8683-induction-deduction.html
 

Salomé

meh
Joined
Sep 25, 2008
Messages
10,527
MBTI Type
INTP
Enneagram
5w4
Instinctual Variant
sx/sp
No doubt. I think it's one of the biggest issues in the subject of cognitive science. I remember having a discussion about this a while back ago with somebody else on this forum who's majoring in that. It's pretty fascinating, if you try and attack it from the mathematical level and not the biological level. There two very different approaches to attacking it.

Do you think one would prove more fruitful to ultimately coming up with an answer. Obviously in the end we would have to attack it from the algebraic end if we are going to possibly try and integrate this with computers, but maybe the the hidden key lies in a different arena and that is why it is such a difficult topic to crack.

Do you have a thread ref?

Everything can (and must?) be reduced to mathematics, ultimately.

Why obviously?
 

Simplexity

New member
Joined
Jul 15, 2008
Messages
1,741
MBTI Type
INTP
Yea I just added the link in the post above.

I was saying that if we want to implement it in the form of technology, it would probably be easiest if it was in computational terms.

The question I had was if maybe focusing on that aspect too much would blind us to some of the ways in which our brain works. Sort of like some of these F v T threads, to a certain extent we can understand the other but there is still a barrier there, we need to hear opinions from multiple sides to form a solid well rounded understanding of the issue.

I actually remember reading somewhere in a book, that they would always try and relate the latest groundbreaking technology to how our brain processed things. For example the Tabula Rasa concept ( I know its not technically for that reason) but when it was originated it seemed like a technologically plausible reason. There's better examples, maybe ten years ago with the outdated processors and computers we were drawing parallels that were related to the then "ground breaking technology"
 

runvardh

にゃん
Joined
Jun 23, 2007
Messages
8,541
MBTI Type
INFP
Enneagram
6w7
Instinctual Variant
sx/so
And past this we can only go into philosophy as science hasn't even gone past this point yet.
 
Top