It is currently Thu Mar 28, 2024 2:23 pm

All times are UTC - 8 hours [ DST ]




Post new topic Reply to topic  [ 157 posts ]  Go to page Previous  1 ... 4, 5, 6, 7, 8  Next
Author Message
Offline
 Post subject: Re: Commonsense Go
Post #121 Posted: Thu May 04, 2017 6:22 pm 
Dies with sente

Posts: 101
Liked others: 24
Was liked: 16
djhbrown wrote:
My bus is going in the wrong direction, because all the passengers want to go the other way.
Of course, i am not one to jump to unwarranted conclusions based on a cursory scan of a couple of pictures - but since everyone around here agrees with me, i must be Right.
Attachment:
q&a.png

If this is the best the best AI can do, then AI is not Going down the tubes, it has already Gone down them, for it is no better than before Dewey invented his decimal classification system, when text-based corpus retrieval by mindless robots (called monks in those days) was just as good then as it is now.


Remove the quotes, try the search again.

Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #122 Posted: Thu May 04, 2017 11:00 pm 
Lives in gote
User avatar

Posts: 392
Liked others: 23
Was liked: 43
Rank: NR
:) ACM had put the quotes in for me, i forget now how i got there, something about subspecialist areas, anyhow, here's what your sensible advice produced after i further modded it to restrict search to abstract so as to omit authors called Go. You will see that the cc search criterion is missing but i can't be fussed to find out how to include it to do a fair comparison.

It's a big improvement, even if ACM doesn't know the difference between Go and go... meanwhile, back at the bike shed, look what a difference a little intelligent planning makes
Attachment:
round2.png
round2.png [ 160.25 KiB | Viewed 10484 times ]
Attachment:
bike2.JPG
bike2.JPG [ 105.39 KiB | Viewed 10484 times ]
Attachment:
me1.png
me1.png [ 125.6 KiB | Viewed 10482 times ]

_________________
i shrink, therefore i swarm

Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #123 Posted: Thu May 04, 2017 11:20 pm 
Lives in gote
User avatar

Posts: 392
Liked others: 23
Was liked: 43
Rank: NR
i rest my case
Attachment:
Screenshot from 2017-05-05 16-16-58.png
Screenshot from 2017-05-05 16-16-58.png [ 1.06 MiB | Viewed 10481 times ]

_________________
i shrink, therefore i swarm

Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #124 Posted: Fri May 05, 2017 6:33 am 
Lives in sente

Posts: 1037
Liked others: 0
Was liked: 180
djhbrown wrote:

But that's not all, for if the Go playing equivalent of Google search is truly better than any living human at Go, which will be proven to be the case in a couple of weeks, then it follows unerringly and unquestionably that Go is a complete waste of time, for as an intellectual task, it requires nothing deeper than mechanical and superficial search by a couple of thousand bots kneejerking their way through a finite forest of improbable moves.


We have a fundamental disagreement, perhaps on the level of "religion"

WHAT do you think "intellect" is? If you want to consider the PROGRAM running on a neural net composed of "a couple thousand mindless bots" how is that FUNDAMENTALLY different from programs running on a neural net composed of "a couple million mindless cells"?

Are you saying things like:
Because the difference is in one case electrical charges and in the other electro-chemical?
<< in the earliest days, there were even hydraulic computers >>. Ultimately a "program"
is something in the realm of mathematics, divorced from the material/physical universe

Because there is something MAGICAL about human intelligence.
<< well I am not going to argue "religion" with you >>

From my point of view, the wonder is that a neural net with JUST a few thousand mindless "cells" can play go as well as it can. But remember, our much larger, more complex neural nes (our human brains) are far more general purpose. At the same time able to play go able to notice that a blue giraffe wandering into the room is an unusual event << but a human in a blus suit something to ignore >>

The fact that I feel that I am "conscious" to me means that "a neural net of sufficient size/complexity can feel itself to be conscious". If it doesn't mean that to you, well and good, but we may not have a common frame of reference to discuss things like this. While I do have my own "religious" beliefs, I do not confuse these with my beliefs about the realms of the material universe or mathematics.

Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #125 Posted: Fri May 05, 2017 2:29 pm 
Lives in gote
User avatar

Posts: 392
Liked others: 23
Was liked: 43
Rank: NR
Mike wrote:
we may not have a common frame of reference to discuss things like this.
i agree.

Mike wrote:
While I do have my own "religious" beliefs, I do not confuse these with my beliefs about the realms of the material universe or mathematics.
the notion of a metaphysical superpower was a subject that puzzled me as a child, which i resolved to my own satisfaction in 1963, when i was 14, on the road to Ramsden Heath, as recounted in "Liam's Book: The Material of the Spirit".
http://lcipm.blogspot.com.au/2013/04/th ... pirit.html

i did not think further about it until 2005 when i put one and one together; the first was a packet of bush tomato plant seeds; and the second was something that Google told me about when i was researching the history of numbers. Together, these two things provided me with an explanation as to why other people believe in metaphysical superpower(s). That explanation is here:
https://www.youtube.com/playlist?list=P ... Q2h2UALN0h

Mike wrote:
If you want to consider the PROGRAM running on a neural net composed of "a couple thousand mindless bots" how is that FUNDAMENTALLY different from programs running on a neural net composed of "a couple million mindless cells"?
there are so many ways in which artificial neural nets are both similar to and different from biological ones that it would take a whole book to describe them. As it happens, there is such a book - it's called "Living Computers: Intelligent Plastic Machines"
http://lcipm.blogspot.com.au/2013/11/li ... astic.html

_________________
i shrink, therefore i swarm

Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #126 Posted: Fri May 05, 2017 3:16 pm 
Lives in gote
User avatar

Posts: 392
Liked others: 23
Was liked: 43
Rank: NR
Mike wrote:
From my point of view, the wonder is that a neural net with JUST a few thousand mindless "cells" can play go as well as it can.
I wondered about that wonder too, ever since she made her dramatic debut in 2015.

I think i have now figured it out, thanks to the DARPA video. Let's see if i can summarise briefly why Alphago is so good at Go:

The game of Go is played within the confines of an artificial microcosm, namely, a universe of just 361 pixels, each of which can have just one of three values.

There is no dark matter in that universe (other than Swim shadows...), nor indeed any energy at all, because the pixels do not move; they simply appear and disappear, according to the whims of the gods that put them there and take them away.

These gods have their own desires, but for the purposes of understanding why Alpha is the Omega of Go, we can safely ignore them, for they are peripheral to the central issue and thus inconsequential.

The universe of Go is minute, a mere tree of 10up170 microphotographs. Some while ago, Japanese theorists noticed that the universe could go on forever in a kind of Kurt Godel way, with true but unprovable theorems about who won, so they put a stop to all that nonsense by modifying the ko rule so that triple kos were a thing of the past, but not of the future.

The tree of Go isn't a proper tree, but one whose branches can grow into one another, producing a net - but, thanks to those Japanese rulemakers, a net for which the arrow of time is steadfastly linear. Dick Whittington can turn, but Theresa May and Go can't go back (not that either of them wants to).

Now, although the tree is not a tree but a net, for the purposes of mathematical analysis, we can transform the net into a tree without losing logical consistency, consistent with Godel's Theorem.

The root of the tree is a cold, dark, place, a yawning empty void of nothingness, with (like our own real universe) no vacuum energy at all - a contemporary myth which will make quite a few people red-faced in a few hundred years' time when it is proven beyond all reasonable doubt that they are talking bullshit because they live in a world of mathematical mythturbation which has nothing even remotely to do with the way the actual thing works - superposition, my foot!.

The leaves of the tree are the ends of the universe, each of which has just one of two possible states: black or white. My guess is that, the Deep South and Brexiteers notwithstanding, it will someday be proved that black is superior to white. Certainly, in my own case, i know that to be unequivocally true, for black people are stronger than white ones like me, because they can withstand the onslaught of B-frequency solar radiation, whereas i now have to go to the quack every so often to have a piece of arm or hand cut out because it has lost its mind and is growing out of control.

Now, because every leaf has a binary value, if we apply Hawking's reasoning and run the clock backwards, we can figure out by linear extrapolation that the beginning of time - when black makes a Big Bang on the board by planting a stone on it - also has a binary value, for there is nothing white can do to stop the inexorable March of Time flying like a banana to the inevitable End of Days.

Time waits for no man, and there are bicycles to repair and bicycles to ride and golfballs to strike and surfboards to fall off of, and George Galloway GG epistles to enjoy, and Dylan Thomases to read, and floors to sweep, and so many things, so little time, so here endeth the Lesson. Tune in next week, same time, same channel, for the next thrilling episode of Agatha Christie's whodunnit.

PS Just heard - read - that AGF are going to do for KGS what [unfinished]. Now that TM has declared war on Europe, i suppose AGF will be launching Go missiles at EGF (kind of an ambiguous acronym) from their colony across the ditch in a vain attempt to reverse the Battle of Hastings. That's a good thing, because it will keep them busy irradiating the Northern Hemisphere, which is soon to become the Southern Hemisphere when the Poles reverse (any day now) and then NZ will take its rightful place at the top of the map like it always said it was and YUK/US will be downvoted so much that it will dissolve into the Land Beyond the Paile, somewhere near Timbuktu. See you On The Beach.
http://www.pottonandburton.co.nz/store/ ... e-down-map

_________________
i shrink, therefore i swarm


Last edited by djhbrown on Fri May 05, 2017 8:19 pm, edited 2 times in total.
Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #127 Posted: Fri May 05, 2017 8:16 pm 
Dies with sente

Posts: 101
Liked others: 24
Was liked: 16
djhbrown wrote:
The game of Go is played within the confines of an artificial microcosm, namely, a universe of just 361 pixels, each of which can have just one of three values.

There is no dark matter in that universe (other than Swim shadows...), nor indeed any energy at all, because the pixels do not move; they simply appear and disappear, according to the whims of the gods that put them there and take them away.

These gods have their own desires, but for the purposes of understanding why Alpha is the Omega of Go, we can safely ignore them, for they are peripheral to the central issue and thus inconsequential.

The universe of Go is minute, a mere tree of 10up170 microphotographs. Some while ago, Japanese theorists noticed that the universe could go on forever in a kind of Kurt Godel way, with true but unprovable theorems about who won, so they put a stop to all that nonsense by modifying the ko rule so that triple kos were a thing of the past, but not of the future.

The tree of Go isn't a proper tree, but one whose branches can grow into one another, producing a net - but, thanks to those Japanese rulemakers, a net for which the arrow of time is steadfastly linear. Dick Whittington can turn, but Theresa May and Go can't go back (not that either of them wants to).

Now, although the tree is not a tree but a net, for the purposes of mathematical analysis, we can transform the net into a tree without losing logical consistency, consistent with Godel's Theorem.

The root of the tree is a cold, dark, place, a yawning empty void of nothingness, with (like our own real universe) no vacuum energy at all - a contemporary myth which will make quite a few people red-faced in a few hundred years' time when it is proven beyond all reasonable doubt that they are talking bullshit because they live in a world of mathematical mythturbation which has nothing even remotely to do with the way the actual thing works - superposition, my foot!.

The leaves of the tree are the ends of the universe, each of which has just one of two possible states: black or white. My guess is that, the Deep South and Brexiteers notwithstanding, it will someday be proved that black is superior to white. Certainly, in my own case, i know that to be unequivocally true, for black people are stronger than white ones like me, because they can withstand the onslaught of B-frequency solar radiation, whereas i now have to go to the quack every so often to have a piece of arm or hand cut out because it has lost its mind and is growing out of control.

Now, because every leaf has a binary value, if we apply Hawking's reasoning and run the clock backwards, we can figure out by linear extrapolation that the beginning of time - when black makes a Big Bang on the board by planting a stone on it - also has a binary value, for there is nothing white can do to stop the inexorable March of Time flying like a banana to the inevitable End of Days.

Time waits for no man, and there are bicycles to repair and bicycles to ride and golfballs to strike and surfboards to fall off of, and George Galloway GG epistles to enjoy, and Dylan Thomases to read, and floors to sweep, and so many things, so little time, so here endeth the Lesson. Tune in next week, same time, same channel, for the next thrilling episode of Agatha Christie's whodunnit.


Do you really have a purpose with these posts, or are you just bored and entertaining yourself? You better implement Swim than waste time this way.

Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #128 Posted: Fri May 05, 2017 8:23 pm 
Lives in gote
User avatar

Posts: 392
Liked others: 23
Was liked: 43
Rank: NR
why do you waste your valuable time reading this rubbish? It's only commonsense that only people interested in commonsense would want to talk about commonsense and draw analogies between Go and real life.

I see you missed the point about why Alphago is so powerful, but i guess that given your name you already know.

That being the case, perhaps you would be so kind as to explain it to us outsiders?

_________________
i shrink, therefore i swarm

Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #129 Posted: Sat May 06, 2017 9:53 am 
Dies with sente

Posts: 101
Liked others: 24
Was liked: 16
djhbrown wrote:
why do you waste your valuable time reading this rubbish? It's only commonsense that only people interested in commonsense would want to talk about commonsense and draw analogies between Go and real life.

I see you missed the point about why Alphago is so powerful, but i guess that given your name you already know.

That being the case, perhaps you would be so kind as to explain it to us outsiders?


Sorry for being so negative - maybe I misunderstood your previous post, I was under the impression that you just write random things that go nowhere.

I am not qualified to explain AlphaGo to you. I do respect systems that work though, like AlphaGo. I believe their success is due to finding a good balance between science and engineering, and being brave to try "crazy things" until it paid off.

Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #130 Posted: Sat May 06, 2017 2:04 pm 
Lives in gote
User avatar

Posts: 392
Liked others: 23
Was liked: 43
Rank: NR
there are too many puns in alhabetty's first sentence, all equally good, that to reply to any one would be to do a disservice to all the others - and so much depth in the second sentence that to even go there would be to distract from the purpose of this post - but i do like the idea of a "balance" between science and engineering :)

last time i wrote something for young children about alphadoggie, the first response was from someone complaining that they didnt understand it, and the second from someone else writing cliffnotes, which was rather nice of them, i thought.

however, i dont expect anyone to be fussed to write cliffnotes on #128, since almost by definition, anyone who could wouldnt have read it, as they are too busy publicising their own ideas and/or selves in more profound yellow presses than a mere grafitti wall.

so i will have to do it myself, if i want it to be done, which i do, so here Go-es:

1. the universe of Go is short - just 361 moves deep, give or take a couple
2. it's not that wide either - at most 361
3. it has no cycles
4. it has a binary ending.

these four characteristics are what makes Alphadog possible, for reasons i will explain. but before doing so, i wish to point out that none of them are characteristics of the real world, which has cycles, and no depth or width boundary as far as anyone knows, except the personal boundary of an individual organism like you or me, which does have significant implications for personal decisionmaking, (as i have learned to my own cost), and many many more branches at each tick of the clock than there are elementary particles in any piece of wood and a couple of bowls of stones.

Of the 4 characteristics, the 4th may be in some ways the most significant, because it's this one that makes neural nets effective.

Back in 1969, Minsky and Papert did some sums on the nets of the day, and figured out that there was something no neural net could ever do: see the hole in a doughnut.

How ironic, for being able to distinguish convex from non-convex is step 1 of every baby Go player's lesson ladder. If you can't see eyes, you can't play Go.

And yet, and yet, although all the backpropagation and layers of convolution in the world don't make any difference - and anyone who says otherwise is either a fool or a liar - so Alphadoggie cannot see eyes either,....

...BUT...

....she can tell black from white, because, as DARPA showed us, that's what neural nets do. The simplest of nets - Perceptrons - drive planes between points scattered in Euclidian space. DCNNs do the same, except that they drive planes between manifold transformations of that space. Whether DARPA's "manifolds" are the manifolds of topology i can't say for sure, it's out of my depth, but i think from his pictures it probably is what Launchbury means.
https://en.wikipedia.org/wiki/Manifold

Nevertheless (again i'm out of my depth here, it really needs a proper mathematician to check my homework), the one discrimination they cannot make; the thing they still cannot do, not unless every input to every convolution contains the entire image, they cannot tell the difference between a sphere and a torus - they cannot tell whether there is a hole in the doughnut.

Le Cunn was on the right track to use DCNN to tell small symbols like the ones you're reading apart, because you can feed the entire image of an alphabettycal character into every prior of every convolution because they're so small.

But a Go board is a bit bigger than a pixellated character, maybe too big

!!BUT!! that's what makes this all so remarkable - Alfie doesn't need to be able to see eyes, because (a) the universe is finite depth - emphasis on the word "depth" and (b) it has only one of two endings, making bifurcation possible.

Goodness, i thought it would take me several postings to explain it all, but i've done it in one go. It's just like the history of religion - i had thought it would be very intricate with lots of crossovers, but it turned out that i could fit the entire history of all the world's religions into one table on one A4 page, because the genealogy is linear (Jesus is another name for Horus = Apollo = Sol Invictus = Stonehenge = the Siberian Father Christmas = the Mayan sun-god = etc etc, and Mary = Isis = Ishtar = Easter = Inanna = the Moon = daughter of the Moon = Venus = the length of the human female menstrual cycle = where babies come from).

Oh, except for explaining why Ke Jie's neural net can't do it too. I will do that tomorrow.
PS and i need to explain why, even though he can't see as well as Alfie, he can, like you and me, and unlike Alfie who can't, he can see eyes. How he does that, how you do that too, is all to do with frogs...
PPS the A4 page is here: https://www.youtube.com/watch?v=GB_Ef1H ... 0h&index=7

_________________
i shrink, therefore i swarm


Last edited by djhbrown on Sat May 06, 2017 3:20 pm, edited 1 time in total.
Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #131 Posted: Sat May 06, 2017 3:18 pm 
Dies with sente

Posts: 101
Liked others: 24
Was liked: 16
djhbrown wrote:
so i will have to do it myself, if i want it to be done, which i do, so here Go-es:

1. the universe of Go is short - just 361 moves deep, give or take a couple
2. it's not that wide either - at most 361
3. it has no cycles
4. it has a binary ending.


Thank you for summarizing your previous post, I see now that you meant some deep things, but I missed them because of all the surrounding noise/jokes.

About "cannot see the hole in the donut" - I am not sure if you mean Minsky's original misunderstanding from the 60s related to the XOR problem with the original perceptron. You must be aware that there are non-linear activation functions that solved that problem long time ago.

Then you extrapolate it to "AlphaGo cannot see eyes" - maybe I don't understand what you mean exactly; it surely knows about eyes very well, judging by how well it plays.

On a more philosophical note: in a sense, one doesn't need to know about eyes in go, since eyes are like a "theorem" in mathematics; all one needs to know are the "axioms" (the rules). Eyes are a consequence of the capture rule, and the fact that people cannot play twice in a row. So maybe you mean that AlphaGo only knows/cares about axioms and not theorems? And that bothers you somehow, why?

Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #132 Posted: Sat May 06, 2017 4:54 pm 
Lives in gote
User avatar

Posts: 392
Liked others: 23
Was liked: 43
Rank: NR
Now look here - you're not supposed to reply so fast, because that's only going to make me want to explain myself, and i've got better things to do, because there are 3 bikes out there all demanding my attention, each with its own special needs, so how any mother can cope with a brood of humanoid bicycles is a total mystery to me, and i bow in all humility to their accomplishment.

maybe xor is solved, but the topological distinction isn't. it isn't. it theoretically cannot be. but i'm not good enough with pencil and paper to draw all the squiggles needed to prove that to the satisfaction of the editor of Ars Mathematica - if you are, i would love to see it written out; it might be easier to do than i think.

Alfabuttery cannot see eyes, period. she cannot tell the difference between "C" and "O". Le Cunn can, because he puts all the corners of the board into the picture - but alfie doesn't.

Ke Jie and you can - but i'm not going to tell you why until tomorrow, because it will take too long and like i said, there are bicycles to feed, and stroke, and i dont want to spend my whole bloody life sitting at a keyboard, expecially as i've now eaten both croissants and am ready to go out and face the day and pay homage to Horus.

But i will say just this: yes, eyes are indeed theorems that can be proven from axioms - as it happens, that's exactly what Swim's colour map does.

But alfabetter doesn't. she can't. BUT she doesn't have to!!!!!!!! Because she can see all the way to the end of time, and tell black from white, and back it up back to the beginning, to know which step to probably take to probably get to where she wants to go. yes, i did miss out all the stuff about Monty-CarloPythin, which i suppose i have to say, but not now, not now.

BTW alfie doesn't bother me - she excites me, she excites my imagination. Or, rather, she did, until the penny dropped, and now that i get it all, she is as dull as dishwater to me and i'm ready to move on to more interesting things, such as putting Kismet's head on top of Nao's body and putting icGo into her mind, to make an emotional Go-playing robot with consciousness and empathy. Kismet is featured in "Common Ground" - my most important movie ever, my legacy to the world for the benefit of future mankind, which only 14 people have ever seen even though it's been there for years.
https://www.youtube.com/watch?v=rIcOpCx ... h&index=10

And there's still the problem of how she can learn, but by following in the footsteps of Isaac Newton, Niels Bohr, and Albert Einstein and forgetting about it, just letting my mind calm itself under a tree, or at the racecourse, or in a tram in (forgot the name of the town, now, i'm sure you know which one i mean - Basel, was it?, you know, the one with the big clock at the end of the street) my subconscious will do its stuff without my help and the answer will spring out like a rabbit from a hat tomorrow morning when (if) i wake up again.

_________________
i shrink, therefore i swarm

Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #133 Posted: Sun May 07, 2017 5:40 am 
Lives in sente

Posts: 1037
Liked others: 0
Was liked: 180
Mistaking the "fundamental difference" that is preventing meaningful discussion.

And actually, the slight diversion by the "religion" issue useful in making this fundamental difference clear.

When you (somehow) jump from what alpha go is/can do to what neural nets (ANY POSSIBLE) are/can do or when you decide the definition of "religion" means metaphysical belief about "superpowers" because you can look at a list of religions of that sort and say "that's what a religion MUST be" you are doing the same sort of thing. Except in the latter case I could point to those few "religions" that while certainly metaphysical, do not have "gods" or at least not a god with any power except existence*. I can't do that with artificial neural nets, none yet in existence large enough or complex enough. But there is no theoretical limit on the size/complexity of an artificial neural net.

Understand? When I say "a neural net COULD in addition to playing go explain why it made certain moves" or "a neural net COULD be conscious" I am not referring to any currently existing artificial neural net (one running on electrical hardware). I can be confident about those statements because I CAN point to an existing neural net able to do those things (it happens to be running on biological hardware) and I trust mathematics where a "program" is an abstract entity independent of any hardware which happen to be running it.

* Of course you might try to wiggle by deciding that your definition determines that things like those branches of Buddhism or for that matter, the pantheism of Spinoza, are not "religion". I would say that they are, also other "not really gods" beliefs of animists or the "kachinas" of the Hopi.

Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #134 Posted: Sun May 07, 2017 7:10 am 
Lives in gote
User avatar

Posts: 392
Liked others: 23
Was liked: 43
Rank: NR
the reason alfie (and JueYi, which i figure to be a clone of her) is God of Go is

(a) because Go has finite depth and ends in a binary result, CNN is good enough. DCNN even better, because the search is better constrained. RL essential for statistically sensible guided search.

+

(b) Monte sampling is good enough for a homogeneous space (ie populated by objects of a single type - in this case, 19x19 pixels)

(c) the more grunt you have, the more you can search, so the better your chances of not missing out on anything good that happens to be within reach. With oodles of doodles, Google can supply a truckload of truckloads of processors to keep the peasants at bay, as can anyone who can dole out 5x$300k from their spare change for a week's holiday without blinking an eye.

(d) Poor old Ke Jie can't compete with that, because for all his brilliance against other H.Saps, he can't read ahead 361 moves, and alfie, although she can't read it all, can read wide enough to read him out of his depth, and her DCNN can eliminate a lot of the nonsense width that no sane bot not losing would contemplate, so her chances of lucking in on a good enough defence to anything he might try are better than his chances of lucking in on one of her weak spots, as Lee Sedol managed to last year, to come back like a zombie from the cemetery.

This, despite the fact that alfie can't see eyes and KJ can. He can, because his eyes work like a frog's, because both his and a frog's eyes are built to the same basic formula.

That formula includes two basic things:
1. a fovea (actually, 2 foveas, one each side of the nose)
2. muscles to move the focus of the fovea around the scene by squishing the lens and rolling the balls

KJ and the frog use their motion detectors to notice changes as their focus shifts, and thereby to detect edges, and thereby to detect eyes.

Swim sees eyes in a more plodding sort of way, by growing a colour-connection map iteratively (the equivalent of a recursive definition of a connected cluster working backwards), and then walking around inside it, noting the internal colour-controlled spots as it goes.

The reason i mention frogs is because Hubel and Wiesel were the first to discover that some neurons attached to the retina of a frog fire only when a line of them notice a change from light to dark (or vice-versa) at the same time, thus telling the frog's brain that an edge is moving in a certain direction.

https://www.researchgate.net/publicatio ... %27s_Brain

They were careful to say that they weren't saying that frog's eyes and human eyes were the same, as it was in those days still unfashionable to cast humans in the same basket as other beasts of the field - and besides, they hadn't dissected any living human eyes (or at least, they didn't say they had... you never know with mad scientists what they do behind closed doors, but no matter how Frankenstinian they are, they're a whole lot less harmless than Bossy Bankers).

Thou shalt have no other God than She who must solemnly be Obeyed, as Rumpole and Basil put it, so it's no wonder there are so many ardent disciples ready to stone anyone that dares take Her name in vain, as recorded in the Life of Brain, and nightly on the Googlebox.

Biographical Note

in 1976, i entered a new world, the world of the Lecturer, a step up
from Assistant Lecturer (another name for donkey worker) and looked
forward to a new life free from the tedious moronities of programming,
which i could delegate to lesser peasants (which in Newcastle in 1988
i actually managed to do, lucking in on two such ******** - Bernie and
Bertrand - who spoke geekspeak well enough to do in 5 minutes what
would have taken me 50 years)

Here's an example of the sort of **** programmers have to put up with
(Tiger used to say that engineers are a breed apart, which is so true,
but they are closer to normal humans than are software engineers, who
are ******* ******).

It occurred when attempting to escape from the clutches of Fedora into
the greener arms of KDE Neon, but unfortunately Fedora's open source
***** have ****** up its configuration so much that it's as
useless as Xubuntu - 10 years ago, both of them used to work, but then
the geeks moved in and ********** it all up.

All because kdenlive, which used to work, no longer does because some
**** thought they knew better, which they didnt.
Code:
[d@localhost ~]$ su
Password:
[root@localhost d]# ./home/d/Downloads/RosaImageWriter/RosaImageWriter
bash: ./home/d/Downloads/RosaImageWriter/RosaImageWriter: No such file
or directory
[root@localhost d]# /home/d/Downloads/RosaImageWriter/RosaImageWriter
No protocol specified
QXcbConnection: Could not connect to display :0
Aborted (core dumped)
[root@localhost d]# **** you
bash: ****: command not found...
Install package 'the****' to provide command '****'? [N/y]

Tempted though i was, i didn't.

Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #135 Posted: Tue May 09, 2017 5:19 am 
Lives in gote
User avatar

Posts: 392
Liked others: 23
Was liked: 43
Rank: NR
the different views of Crazy Sensei (CS) and Swim of the same position, taken from the game discussed in
viewtopic.php?p=217844#p217844
Attachment:
csmap.png
csmap.png [ 99.83 KiB | Viewed 10077 times ]
The image on the left is CS's Territory view; presumably the sizes of territory points reflect CS's confidence in their becoming territory, presumably derived from lookahead simulations.

On the right, Swim's preliminary perceptions derive purely from her colour and shadow map algorithms, without any lookahead. Swim's view is hand-drawn on a picture of CS's policy net move recommndations; presumably the bigger the red, the more strongly CS recommends it.

Both views infer the white stone on the upper left is dead. There are other similarities, but some big differences. For example, CS gives white about 34 points of territory at the top, but Swim sees only one eye point (but a large eyespace).

Swim sees lots of white shadow; if you add up all the territory and shadow points for each side, i think you will find that Swim puts white in the lead.

Swim's further analysis and her own move suggestions have not been simulated, as i've put her to work on a different problem:
viewtopic.php?p=219122#p219122

If anyone reading this is interested in their own analytical capabilities, they might like to find their own answers to Kirby's question, so they will have something to compare with Swims analysis when it's produced, in a few days' time.

corrigendum: the shadow map in this example is slightly wrong, because i didn't follow my own algorithm properly. the border of the white shadows at the top and left is too straight; they should be a little more rounded; i had overlooked some propagations, having confused myself in an attempt at drawing pretty shadows. But that error doesn't make any substantive difference to the bigger picture.

_________________
i shrink, therefore i swarm


Last edited by djhbrown on Thu May 11, 2017 4:11 pm, edited 1 time in total.
Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #136 Posted: Thu May 11, 2017 3:28 am 
Judan

Posts: 6725
Location: Cambridge, UK
Liked others: 436
Was liked: 3719
Rank: UK 4 dan
KGS: Uberdude 4d
OGS: Uberdude 7d
From Swim's Reviews thread:

djhbrown, so I/we can be clear what is going on here, could you please answer a few questions:
1) Are Swim, Commonsense Go, icGo etc the same thing?
2) Are you applying your Swim ideas by hand or with a computer? If the former how are you implementing functions like weak() in a way that means your human Go skills and biases do not come into play such that this becomes more of a "human Go player djhbrown reviews your game with the help of a territory/shadow map and some rules/heuristics", rather than "an algorithm that could be implemented in a computer to play Go reviews your game"?

Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #137 Posted: Thu May 11, 2017 4:56 am 
Lives in gote
User avatar

Posts: 392
Liked others: 23
Was liked: 43
Rank: NR
All Uberdude's questions are answered in icGo documentation https://sites.google.com/site/djhbrown2/icgo
which is from time to time updated with new and/or corrected information, including the meat of Swim's Reviews

But since he was kind enough to repost his questions here, as asked, to protect the coherence of the other thread, i answer them here too:


1) Are Swim, Commonsense Go, icGo etc the same thing?

Commonsense Go -> Swim -> icGo

2) Are you applying your Swim ideas by hand or with a computer?

by hand. following the algorithms. the only software implementation that i know of is pnprog's Gomap


If the former how are you implementing functions like weak() in a way that means your human Go skills and biases do not come into play such that this becomes more of a "human Go player djhbrown reviews your game with the help of a territory/shadow map and some rules/heuristics", rather than "an algorithm that could be implemented in a computer to play Go reviews your game"?

weak is a qualitative value, not a function. you can think of it as just a string of letters (or a logical value). it is defined on p32 of running draft of icGo documentation https://sites.google.com/site/djhbrown2/cg.odt, unchanged from original paper on CG. There is a predicate weak(group) in the value function, but all it does is return True if group.status = weak.

there was a time when qualitative computation was a hot topic in AI, but the moneybags of first IBM and now Google have put paid to that for a while longer - until they finally find out that they're going the wrong way off a Parkinsonian cliff https://en.wikipedia.org/wiki/Parkinson%27s_law :) Except in Go and Jeopardy, that is, where they work just fine. I personally think they're going to come a cropper when it comes to medical diagnosis (let alone meaningful dialogue in natural language), but that's Another Story for another forum or two.

the game Swim is reviewing right now in Swim's Reviews is between a 3d and a 2d. i am 5k on Pandanet (see my avatar). does that answer your question?:)

the original paper on Commonsense Go reviewed a move in the Alphago - Lee Sedol match. i'm not quite as strong as Lee Sedol...

The purpose of Swim's reviews is twofold: for the requester, it will produce some insights that will hopefully be valuable to him/her; and for me, it will show me where i need to modify and/or augment what is already defined. there have been a few minor changes to the definition of the colouring procedures, but their basic principle is unchanged. there are some places in the methods which are not fully defined yet, and i anticipate that doing reviews where i do not know the "answer" beforehand will highlight to me what needs to change.

for me, producing colour and shadow maps is very tedious, but i'm starting to get the hang of it - mainly, by refraining from leaping ahead locally and being more systematic. i still haven't found a way to hand-draw pretty shadows, something that a computer would find a piece of cake. Just need a programmer...

i'm never going to try to code up icGo myself; i'm more interested to see if i can apply https://papers.ssrn.com/sol3/papers.cfm ... id=1340783 and https://ssrn.com/abstract=2205530 to icGo


_________________
i shrink, therefore i swarm


Last edited by djhbrown on Fri May 12, 2017 4:30 am, edited 1 time in total.
Top
 Profile  
 
Offline
 Post subject: Re: Commonsense Go
Post #138 Posted: Thu May 11, 2017 6:43 am 
Lives in gote
User avatar

Posts: 392
Liked others: 23
Was liked: 43
Rank: NR
"how are you implementing functions like weak() in a way that means your human Go skills and biases do not come into play..?"

i'm not :) my biases come into play at each and every step of Swim's algorithms, because i wrote them!

but i'm not Swim. she is much stronger than me, of that i am 100% sure. because, for a start, she doesn't miss atari like i did in a game today and lost a game i had been way ahead in for about 50 moves.

an honourable opponent would have resigned long before it happened, but most of my opponents are anything but honourable :)

but more significant than that, is that Swim is already teaching me some things i didn't know.

the first one turned up in some test examples pnprog had constructed to test Gomap. it relates to a standard keima kakari to hoshi joseki, where after a tobi response, the kakari is followed up by a slide under the hoshi, met with a kosumi block, and then the kakari-er jumps back along the side to make a base.

except, that wasn't the actual pattern tested, because pnprog had the stones on the 4th and 3rd lines instead of the third and second,

what turned up - to my surprise - was that the colour map spread right along and under the extension, all because of the slide underneath at the other end. without that slide, that doesnt happpen. im describing it rather than showing it because it's easier for me to type than draw but you can construct it for yourself and follow the colour map steps and see what happens.

then compare it with a 4th line 2-space extension without the slide (you know, the one Alfie loves). the difference is remarkable.

the other example is in Kirby's game, where Swim reckons black has not just influence, but a ton of potential territory right in the middle.

this is totally counterintuitive to me, as all the books i read and believed 40 years ago said "don't use thickness to make territory".

and i still believe those old books - because, even if white neutralises it in the middle, black will still get it instead on the other side of white's reducing moves.

So both Swim and the books are right, even though they say opposite things!

And the implication of this is that, in Kirby's game, for example, white got nowhere near as much as black from their exchanges at the top, which is why he's already lost the game, at move 33!

that last bit is my opinion, not Swim's. so although i mention it here, i won't mention it there, because Swim's review is only Swim's review, not my own.

and ... try this for size: the huge value of centre-facing walls may well be Alphago's opinion too! she plays lots of moves aimed at the centre that lots of people have remarked upon. maybe Swim can explain why Alphago is right, even though Alfie doesn't know what she's doing! :) Swim already did it for that move 37 that the MIT journo (aka Deep Mind PR agent) bangs on about in that article thinly pretending to be about AI and language, which reads like it came straight from the pen of Demis himself. Footballers have ghost writers, here it looks like it's the other way around :)

_________________
i shrink, therefore i swarm

Top
 Profile  
 
Offline
 Post subject: fightback
Post #139 Posted: Sat May 13, 2017 3:47 pm 
Lives in gote
User avatar

Posts: 392
Liked others: 23
Was liked: 43
Rank: NR
what a difference a word makes!
Code:
let lead =  value (me) - value (you);
if lead + chance > 0 then playsafe   
                     else fightback
of course, just changing one word makes no difference at all if the meaning is unchanged, which in Swim's case it isn't.

but it makes all the difference in the world to the way in which it's interpreted by a reader.

in "The Art of War", Sun Tsu advises that you should never start a battle you cannot win.

if Go players were to take Sun Tsu's advice, they would never start a game in the first place, because you can't guarantee to win every time.

As a child of the 60s, i shared the dreams of Martin Luther King and John Lennon, and still dream on, although the last 50 years have made me more cynical than ever that it could ever happen; there are just too many people making too much money out of persuading other people to fight wars.

Up to now, Swim has looked at close games in which playing safe was the (i insist) only sensible option. And two games (Alfie vs Lee and Andrew vs Nick) in which Swim's player was already miles ahead** and just needed to protect his investment. Only in JueYi's New Move did i consider how to go about invasion, drawing upon Haylee's thrashing of Nick in an earlier episode.

Attack is the best means of defence, so Swim's methods for playing safe involve doing things like making leaning attacks.

But in the Kirby game - the first of Swim's reviews - Swim is presented with a new challenge: how to fight back.

This is the least well-developed side of Swim, partly because my own prediliction for running away from conflict, rather than confronting it, had placed my focus upon honte strategy.

But honte's not going to work for Kirby.

all i know about attacking is what i vaguely remember from reading Kato's "Attack and Kill" and now wish i hadn't given all my Go books away in 1983.

i reckon that most of you enjoy nothing more than a full-on blood and guts scrap, so maybe you can help me out by providing a precis of your fighting strategy that i can translate into algorithms for Swim.

Any offers?

ps
I'd like to change the rules of Go to this: both players succeed in co-producing a beautiful drawing when neither kills the other and the territory is roughly equally shared; otherwise they both lose.


pps ** no he wasn't! - Alfie would have been behind if all Lee's groups lived, so the original CG paper already had the methods that Swim needs to help Kirby. I suppose gaga is when you try to solve a problem you solved a year ago...

_________________
i shrink, therefore i swarm


Last edited by djhbrown on Sun May 14, 2017 5:13 am, edited 1 time in total.
Top
 Profile  
 
Offline
 Post subject: Re: fightback
Post #140 Posted: Sat May 13, 2017 10:03 pm 
Dies with sente

Posts: 101
Liked others: 24
Was liked: 16
djhbrown wrote:
i reckon that most of you enjoy nothing more than a full-on blood and guts scrap, so maybe you can help me out by providing a precis of your fighting strategy that i can translate into algorithms for Swim.
Any offers?


Are you sure this is the way to apply AI to go?

Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 157 posts ]  Go to page Previous  1 ... 4, 5, 6, 7, 8  Next

All times are UTC - 8 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group