Contemplation of the naval
-
Sitting here in Samoa (my home) after just finishing a rather stressful job and decided to just let my brain wander. Why can't computers forsee and challenge any human action within a given space ? Chess for instance is very much a controlled environment yet people have, and will continue to, beat computers at it. Is it programmer laziness (or time restrictions) or some intangible quality we posses as humans that make us so difficult to predict ? Or do i just need a beer and some sunshine ? Life in the tropics isn't all it's cracked up to be. There's more beer than the brochures would have you believe
-
Sitting here in Samoa (my home) after just finishing a rather stressful job and decided to just let my brain wander. Why can't computers forsee and challenge any human action within a given space ? Chess for instance is very much a controlled environment yet people have, and will continue to, beat computers at it. Is it programmer laziness (or time restrictions) or some intangible quality we posses as humans that make us so difficult to predict ? Or do i just need a beer and some sunshine ? Life in the tropics isn't all it's cracked up to be. There's more beer than the brochures would have you believe
From what little I know about artificial intelligence - the biggest factor limiting AI is that most algorithms involved are exponential... which means they are prohibitively slow. So you typically have to make up and use heuristics all over the place.. kind of like intelligent guessing. So it's still very much a work in progress, and a lot of it is emperical. I'm sure there are lots of other things too. If your nose runs and your feet smell, then you're built upsie down.
-
From what little I know about artificial intelligence - the biggest factor limiting AI is that most algorithms involved are exponential... which means they are prohibitively slow. So you typically have to make up and use heuristics all over the place.. kind of like intelligent guessing. So it's still very much a work in progress, and a lot of it is emperical. I'm sure there are lots of other things too. If your nose runs and your feet smell, then you're built upsie down.
I'm sure there are lots of other things too. You've never been to samoa have you Navin ??
-
Sitting here in Samoa (my home) after just finishing a rather stressful job and decided to just let my brain wander. Why can't computers forsee and challenge any human action within a given space ? Chess for instance is very much a controlled environment yet people have, and will continue to, beat computers at it. Is it programmer laziness (or time restrictions) or some intangible quality we posses as humans that make us so difficult to predict ? Or do i just need a beer and some sunshine ? Life in the tropics isn't all it's cracked up to be. There's more beer than the brochures would have you believe
Andrew Bleakley wrote: make us so difficult to predict ? Maybe they should consider using random memory address values to simulate misfired neurons, and epiphanies, etc.
"Manifest plainness, embrace simplicity,
reduce selfishness, have few desires."
-- Lao TzuBW
-
Sitting here in Samoa (my home) after just finishing a rather stressful job and decided to just let my brain wander. Why can't computers forsee and challenge any human action within a given space ? Chess for instance is very much a controlled environment yet people have, and will continue to, beat computers at it. Is it programmer laziness (or time restrictions) or some intangible quality we posses as humans that make us so difficult to predict ? Or do i just need a beer and some sunshine ? Life in the tropics isn't all it's cracked up to be. There's more beer than the brochures would have you believe
Andrew Bleakley wrote: Why can't computers forsee and challenge any human action within a given space ? The brain is a highly effecient pattern matching processor. No computer can touch it yet in that capability. Marc Latest AAL Article My blog Join my forum!
-
I'm sure there are lots of other things too. You've never been to samoa have you Navin ??
Hey, I've been to South Auckland ! Which side of the dateline are you ? Regardz Colin J Davies
*** WARNING *
This could be addictive
**The minion's version of "Catch :bob: "It's a real shame that people as stupid as you can work out how to use a computer. said by Christian Graus in the Soapbox
-
Andrew Bleakley wrote: Why can't computers forsee and challenge any human action within a given space ? The brain is a highly effecient pattern matching processor. No computer can touch it yet in that capability. Marc Latest AAL Article My blog Join my forum!
Marc Clifton wrote: The brain is a highly effecient pattern matching processor. Maybe "thinking" is pattern matching followed by learning new patterns. Regardz Colin J Davies
*** WARNING *
This could be addictive
**The minion's version of "Catch :bob: "It's a real shame that people as stupid as you can work out how to use a computer. said by Christian Graus in the Soapbox
-
Hey, I've been to South Auckland ! Which side of the dateline are you ? Regardz Colin J Davies
*** WARNING *
This could be addictive
**The minion's version of "Catch :bob: "It's a real shame that people as stupid as you can work out how to use a computer. said by Christian Graus in the Soapbox
The other side from New Zealand (just). Same side as USA. South Auckland, from what I hear that is just a holding cell for all the Samoan's that have escaped the island.
-
I'm sure there are lots of other things too. You've never been to samoa have you Navin ??
Samoa, yeah, isn't that in New Jersey? :-D Are you talking about American Samoa? Never been there, but maybe someday... I'd like to take an island-hopping trip thru the Pacific and hit Hawaii, Guam, Samoa, and maybe some others. If your nose runs and your feet smell, then you're built upsie down.
-
Samoa, yeah, isn't that in New Jersey? :-D Are you talking about American Samoa? Never been there, but maybe someday... I'd like to take an island-hopping trip thru the Pacific and hit Hawaii, Guam, Samoa, and maybe some others. If your nose runs and your feet smell, then you're built upsie down.
No Samoa proper. The jewel of the Pacific. Definitely do the island hop it is a great way to get married.
-
Andrew Bleakley wrote: Why can't computers forsee and challenge any human action within a given space ? The brain is a highly effecient pattern matching processor. No computer can touch it yet in that capability. Marc Latest AAL Article My blog Join my forum!
Interestingly enough... Most approaches to Chess AI have been in starting with memorization of basic openings, then predicting every possible move using a mini-max or ab pruning tree. Such an approach is (in theory) easy enough in chess because there are a finite number of pieces that can move in a very limited number of ways. The algorithms take a LONG time to compute, but they are "realistically computable". That is why we can develop software/hardware systems that can, on average, beat any normal player, and with enough horsepower, even beat most masters. The game of Go, on the other hand, has a nearly infinite number of possible moves that can be played in a single game, and since the rules of placement are so loose, standard tree-prediction methods are not "realistically computable" (it would take several orders of magnitude longer to compute a single Go move using a chess-like algorithm). All publicly-accessible (meaning real-time playable) Go AI's can be readily beaten by even top-rated amateurs. Go seems to be more of an intuitive, pattern based system. Top players can see patterns developing on the board before they are even there. So, appropriately enough, many attempts at creating a master level Go computer AI system rely on pattern matching principles (one successful program uses a complicated neural net w/an accompanying influence map to plot moves). According to my research, no one has created a Go program that can compete with professional players as of yet, however... Jeremy "Learning the Ways of Go" Kimball
-
The other side from New Zealand (just). Same side as USA. South Auckland, from what I hear that is just a holding cell for all the Samoan's that have escaped the island.
Ah, you are one of yesterdays people. :-) Yeah, South Auckland has lots of Samoans, and other Pacific people. The warfare between them gets out of hand at times, as you can imagine. They seem to either be be super devout religeous types, or super violent. (stereotypes} I think a lot of the poorer Samoans that arrived here as stowaways have had a hard time adjusting. "And I do mean stowaways" Anyhow if it wasn't for them our Rugby teams would be crap. :-) Regardz Colin J Davies
*** WARNING *
This could be addictive
**The minion's version of "Catch :bob: "It's a real shame that people as stupid as you can work out how to use a computer. said by Christian Graus in the Soapbox
-
Interestingly enough... Most approaches to Chess AI have been in starting with memorization of basic openings, then predicting every possible move using a mini-max or ab pruning tree. Such an approach is (in theory) easy enough in chess because there are a finite number of pieces that can move in a very limited number of ways. The algorithms take a LONG time to compute, but they are "realistically computable". That is why we can develop software/hardware systems that can, on average, beat any normal player, and with enough horsepower, even beat most masters. The game of Go, on the other hand, has a nearly infinite number of possible moves that can be played in a single game, and since the rules of placement are so loose, standard tree-prediction methods are not "realistically computable" (it would take several orders of magnitude longer to compute a single Go move using a chess-like algorithm). All publicly-accessible (meaning real-time playable) Go AI's can be readily beaten by even top-rated amateurs. Go seems to be more of an intuitive, pattern based system. Top players can see patterns developing on the board before they are even there. So, appropriately enough, many attempts at creating a master level Go computer AI system rely on pattern matching principles (one successful program uses a complicated neural net w/an accompanying influence map to plot moves). According to my research, no one has created a Go program that can compete with professional players as of yet, however... Jeremy "Learning the Ways of Go" Kimball
On that note about Chess, I have often though it would be possible to decompile and map chess. If you startaed from the end game and found as many possiblities that exist that work them one move backward to determine what the next move should be, then continue until every possible move was mapped from the start game to a termination. Of course this would take lots of grunt, but in the future it may be possible. Regardz Colin J Davies
*** WARNING *
This could be addictive
**The minion's version of "Catch :bob: "It's a real shame that people as stupid as you can work out how to use a computer. said by Christian Graus in the Soapbox
-
On that note about Chess, I have often though it would be possible to decompile and map chess. If you startaed from the end game and found as many possiblities that exist that work them one move backward to determine what the next move should be, then continue until every possible move was mapped from the start game to a termination. Of course this would take lots of grunt, but in the future it may be possible. Regardz Colin J Davies
*** WARNING *
This could be addictive
**The minion's version of "Catch :bob: "It's a real shame that people as stupid as you can work out how to use a computer. said by Christian Graus in the Soapbox
Interesting thought... It WOULD take a huge amount of grunting. I did something very similar for tic-tac-toe (naughts for you non-americans), by mapping out every single possible game state. Huge amount of data, but in theory, you could never beat a system using such data... It feels *icky* somehow, though, to use such a system...almost like cheating... Jeremy
-
Sitting here in Samoa (my home) after just finishing a rather stressful job and decided to just let my brain wander. Why can't computers forsee and challenge any human action within a given space ? Chess for instance is very much a controlled environment yet people have, and will continue to, beat computers at it. Is it programmer laziness (or time restrictions) or some intangible quality we posses as humans that make us so difficult to predict ? Or do i just need a beer and some sunshine ? Life in the tropics isn't all it's cracked up to be. There's more beer than the brochures would have you believe
Andrew Bleakley wrote: some intangible quality we posses as humans that make us so difficult to predict ? I think it is creativity that make us humans very difficult to predict. We can apply skills that we have learned in one area of life to something totally new. Andrew Bleakley wrote: Life in the tropics isn't all it's cracked up to be. There's more beer than the brochures would have you believe :laugh: I have the same experience jhaga --------------------------------- I have discovered that all human evil comes from this, man's being unable to sit still in a room. Blaise Pascal (1623 - 1662)
-
Marc Clifton wrote: The brain is a highly effecient pattern matching processor. Maybe "thinking" is pattern matching followed by learning new patterns. Regardz Colin J Davies
*** WARNING *
This could be addictive
**The minion's version of "Catch :bob: "It's a real shame that people as stupid as you can work out how to use a computer. said by Christian Graus in the Soapbox
Colin Davies wrote: Maybe "thinking" is pattern matching followed by learning new patterns. A lot of it is. Voice recognition, singling out a conversation in a crowded room, identification of objects, driving, all sorts of stuff. But there are other realms of thinking too. The expression of concepts and deductive reasoning, for example. Marc Latest AAL Article My blog Join my forum!
-
Interestingly enough... Most approaches to Chess AI have been in starting with memorization of basic openings, then predicting every possible move using a mini-max or ab pruning tree. Such an approach is (in theory) easy enough in chess because there are a finite number of pieces that can move in a very limited number of ways. The algorithms take a LONG time to compute, but they are "realistically computable". That is why we can develop software/hardware systems that can, on average, beat any normal player, and with enough horsepower, even beat most masters. The game of Go, on the other hand, has a nearly infinite number of possible moves that can be played in a single game, and since the rules of placement are so loose, standard tree-prediction methods are not "realistically computable" (it would take several orders of magnitude longer to compute a single Go move using a chess-like algorithm). All publicly-accessible (meaning real-time playable) Go AI's can be readily beaten by even top-rated amateurs. Go seems to be more of an intuitive, pattern based system. Top players can see patterns developing on the board before they are even there. So, appropriately enough, many attempts at creating a master level Go computer AI system rely on pattern matching principles (one successful program uses a complicated neural net w/an accompanying influence map to plot moves). According to my research, no one has created a Go program that can compete with professional players as of yet, however... Jeremy "Learning the Ways of Go" Kimball
Very interesting. I was contracted a few years ago to analyze redundancy ring networks for satellites. Everyone had been trying to analyze the problem by optimizing search algorithms that essentially walked through every switch state. For a large ring, say 16 actives and 8 redundancies (backups), two problems exist--one, the analysis time is measured in years, second, the volume of information (in the trillions of path permutations) is pretty much useless. I ended up writing some algorithms that looked at connectivity patterns. Interestingly, this resulted in two things--I could tell the engineer where failures would occur (which is what they wanted to know to begin with) in about 10 seconds, and I also established a set of rules so that the program could generate optimal rings (at least from the network switching perspective--issues such as power consumption and mass were not handled!). So, by teaching myself how to look at the problem from the perspective of patterns, and then figuring out the rules for those patterns (which is really hard, BTW), I was able to do so really cool stuff! Marc Latest AAL Article My blog Join my forum!