In the latest Go World (Sept 2018), O Meien has revealed more thoughts on why humans fail against AI.

He says pros are trying new AI-influenced moves and commentators on TV games are very fond of alluding to AI moves, but this doesn't mean they understand anything yet. Rather they are doing this in the hope of discovering another way up the mountain.

But one thing seems certain already and that is that humans think locally whereas AI takes account of the whole board.

O's interesting take on this is that humans have become fixated on the size of a move. We need some sort of measure to help us navigate through a game, and size of a move has proven to be an incredibly useful measure. Apart from different ways of counting value directly, pros have even established an array of tricks that help them in their quest for a move's value, such as miai and tewari.

But evaluating the size of a move has major drawbacks. For one thing it is based mostly on assessment of the difference of "if I play there or he plays there." It is therefore a very local thing and it is easy to feel that a move is bigger simply because it is in a busier area (e.g. "the opponent's vital point is my vital point"). In the fuseki, a wariuchi splitting move is often seen as common sense for a similar reason, but AI has shown a quiet move elsewhere can be better.

But focusing on the size of move can have more pernicious effects. O says the reason the early shoulder hit so beloved of AI programs has been a blind spot for pros is because the pro just assumes that as defending against is a gain territorially it must be bad. Pros have therefore never even considered such moves as candidate moves.

There is, of course, another measure which pros have labelled 'intuition' or 'feeling' and this does seem to have more of a full-board feel to it. But how do we learn to trust it?

He gives a simple example to ponder on. Take the very well known joseki where Black begins with komoku (c4) but we vary the order of moves (i.e. a version of tewari). So, White plays against that 1. f3, Black 2. d3, White 3. e4, Black 4. e3, White 5. f5, Black 6. c6.

In this order of moves, we can see that the exchange of Black 4 for White 5 is bad, but so is the exchange of White 3 for Black 6. The assumption so far has been that the stupidities cancel out. But in terms of the whole board, which move was more stupid: White 3 or Black 4? It seems as if the task for modern pros is to find a way to make that sort of assessment rather than focusing on the size of a move. That may then lead to playing shoulder hits ourselves.

O's whole text is quite long and has several specific examples. I have latched on to what interested me most, so if you want the more rounded version you will need to look up his text. But I think what I have posted here is enough to indicate how and why pro thinking may evolve away from the size of moves.

Move size (value) need not be abandoned. Quite contrarily, its current concept can be extended to include all aspects incl. value of global interaction, which we must study to measure numerically.

and as RJ is here too, I'll repeat what I said in a recent thread



Is no mention made of white 1, the distant approach? (and I presume f5 meant f4). It's very slack territorally (yes, the same could be said of early shoulder hits ), and although the human knowledge is that it can be good in situations like Kobayashi opening where you don't want to be pincered but emphasise the side (with the latent threat of getting a more efficient corner with attach if they insist on pincering), I expect AIs would have a dismal view of it in most situations strong humans have played it. Without even checking Elf, I predict it will say it's pretty bad even with other corners occupied by positions that aren't begging for an approach/invasion.

I take it you mean the Nihon Kiin magazine.



As I have said, I think that will happen fairly soon, as youngsters who learn from AI come to the fore.



That seems a little broad.



Well certainly the size of a move, even as estimated by Ishida Yoshio, nicknamed "The Computer", has been typically undervalued in the opening. The two moves of an enclosure together have gained 25 points plus, yet books on evaluating positions count territory of less than half that value. It seems to me that pros have persistently undervalued influence. (They may have overvalued it in the New Fuseki, but they have retreated from that.)



It need not be. In fact, my heuristic of comparing two plays by treating them as miai is good as long as neither is sente and they do not interact too much. The miai can be quite distant from each other (in fact, that is preferable), so that the comparison is not local.




Speaking for myself, I have been surprised at how often the AI bots have played in the busy areas. I guess I tenuki too much.



Interestingly, in his thinking about 21st century go, Go Seigen downgraded the wariuchi, although not always consistently.



Again, this seems to me like undervaluing influence in estimating the size of a play.



Consulting the AI bots will be a big help, although it may be hard for an old dog to learn new tricks. Just this morning I was watching a go video that said that you can't learn direction of play from Leela Zero. Au contraire, IMHO. Right now, the best teachers of direction of play are Leela Zero, Elf, AlphaGo, and other top AI bots. It may not be direction of play as I learned it years ago, but you have to discard old ways of thinking.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

As usual, I'm just trying to provoke discussion, but I do have a gut feeling about the above that suggests it could reward more thought.

I am probably the one to blame for the phrase "direction of play." I did not invent it, but propagated it by translating Kajiwara's book. It actually came from Stuart Dowsey at the London Go Centre. When I was working on the book, we were discussing possible titles and for reasons I can't fully remember, I went with Stuart's suggestion. I have regretted it mildly ever since.

Now, in the light of AI's success, I regret it even more.

The reason for my regret is that the Japanese literally means "direction of a stone", but hidden behind that for western players who don't know Japanese is a range of meanings to do with the fact that Japanese does not normally mark nouns as singular or plural. The possible meanings are therefore direction of a stone, direction of stones, directions of a stone, directions of stones. But we have to add on to that the fact that the meaning "stones" is also the normal way to refer to "group" (or groups).

"Direction of play" cuts through that. However it also loses something. I suspect what has happened over the years is that direction of play in the west (and possibly in Japan) has been seen as a vector, i.e. it has magnitude as well as direction. Since we are predisposed to think of move size, as O Meien has pointed out, we run the risk of using direction of play as yet another support in our quest for move size. But AI has shown we should be emphasising the "direction" of the vector much more than the "magnitude". And on top of all that, we should be remembering more that it is directionS we need to think about so that we focus more on the whole board instead of magnitude, which limits our focus too much to what is local.

Studying direction of play is going to be important from now on, but the first step would be a change of name to one that reminds us of everything that is going on behind the Japanese original 石の方向. Any good suggestions?

As a thought experiment on actual study of the concept, could we try overemphasising the direction part of the vector? For example, if we draw lines of the AI's preferred move from each stone/group to other salient points, do interesting patterns emerge? Or interesting differences from lines drawn from our own choice of move? Wilcox used this idea (though Miyamoto preceded him) in a simplistic way with his sector lines. May this idea needs re-visiting?

It doesn't undermine the point you are making, but my memory of what Ishida mainly said is not quite that.

First he stressed he was talking about big points only. Then I think he said that any big point in the opening (first move in a corner or on the side, or shimari or kakari) is worth not quite 20 points, but if the move has a follow-up (or denies the opponent one) it is worth a little more, and he accordingly distinguishes Super Big Points (23-24) points, Class I (22-21 points) and Class II (20-19). If a move is worth 25 points or more it is classed as an urgent point rather than a big point.

Personally I never really understood what these numbers meant or how they were reached, but I'm pretty sure they are specifically to do with big points. What I also remember clearly some (?)40+ years later is that when I first opened his book and saw moves labelled as 20+ points I thought I had found gold in them thar hills. For me, at least, it turned out to be fool's gold.

Ishida also had a view on the value of bad moves. A minor bad move loses 2-3 points. A bad move loses 5 points. A very bad move loses 10 points. Moves that lose 10 points are few and far between. He obviously hadn't seen my games.

Maybe stone's direction or stone direction. Putting the word stone first seems to go some way in producing a similar linguistic effect in English as described above in Japanese.

Or do away with direction altogether and choose between terms such as 'compass of groups' (groups imply single stones in it's definition), 'tide of stones' (think of stones as a wave), 'electro-magnetic movement' (in practical use, the direction of the electromagnet, being prerequisite for function, is more important than the charge it receives), and so on...

_________________
On Go proverbs:
"A fine Gotation is a diamond in the hand of a dan of wit and a pebble in the hand of a kyu" —Joseph Raux misquoted.

"Stone's direction" doesn't attend to the singular/plural problem. We must be careful not to look at one stone/group at a time.

With tongue in cheek, I have considered "lithic radiation" Values can be measured in lithocuries.

I feel this is moving toward the right direction.

Winrates. Maybe we can borrow from quantum terms.
Area/region of interest.
Key area/region.

( Is the temperature thing still good in the age of bots ? )

Reminds me a of a lecture Ko Joyeon 2p gave at the London Open a few years ago. It was all about "lines", and she didn't really manage to explain it very well so I don't remember much, but it did seem to be an interesting and novel approach.

(and also djhbrown! though his approach seemed to be if you draw enough lines on the board then of course a move decided to be good by another approach to go playing will be on some of them, but so will lots of bad moves)

IMO, direction of play is fine. Since the concept really can only be explained by putting stones on the board, I don't think we can ask too much of language alone. I often talk of development, which implies "of one's stones", and at least overlaps with ishi no houkou.

Later I'll post some diagrams from the AlphaGo Teaching tool.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

I loved this article, and I feel that it's highly relevant.

_{I read this one in the terrific short story collection Stories of Your Life and Others. One of the stories in that collection was the basis for the film Arrival. I really enjoyed that movie, but it's not even my favorite story in this collection.}

I agree with Bill that we can't expect too much translating go to English - especially if we go through the intermediate step of Japanese. It seems that in order to find a better term than the already good "direction of play," one must have a pretty good understanding of what is meant as fas as stones are concerned. Since I have lent my copy of the book to a friend, I can't refer to that, so I'll just say that I like both development and movement. Perhaps one could avoid the singular/plural problem by avoiding stones altogether with something like "positional development" or "directional movement in go."

_________________
Patience, grasshopper.

My view is kind of...

Computers have been naturally better than humans at certain activities for a long time. Iteration is a simple example. A computer can multiply two numbers a million times in a row in under a second. For a human to do the same math, it'd take a really long time.

More relevant to games, computers are also really good at search. Like in chess, a computer program can be pretty strong by minimax with some simple heuristic - by the computer's nature, it can hold more in memory than humans at any given time.

With go, for a long time, things were different. Even though computers have various strengths compared to people, human intuition, go theory, etc., could overcome the natural advantages that computers have with search, etc.

But with deep learning, computers caught up on the intuition part. Before, even with natural advantages computers have with search, memory, and stuff like that, human theory was superior so we still had a chance.

Once computers surpassed a fraction of the intuition ability that humans had, we don't have a chance.

It's not because computer theory or intuition is necessarily better than that of humans - it just has to be good enough so computers can surpass us at what computers are good at.

For example, if a computer had exactly my evaluation ability and exactly my go strategy, and exactly my go intuition, it would win. Why? Because it can store more board states in memory than I can.

It's like competing with a calculator to do multiplication. We could both have perfect theory and understanding. But computer will win simply because it has more efficient resources for computing something it understands.

I don't necessarily think that I can learn much math theory from a calculator. But it's for sure better at solving arithmetic. And good for checking my work I guess.

_________________
be immersed

Yeah, some of what Kirby said matches my thoughts.

Having played with Go AI and neural net training over the last 8 months now (as a software developer, in my spare time I've been using Go as an excuse to gain experience in the modern wave of machine learning developments that have taken place in the last 5 years or so), I've played quite a lot now with the raw position-by-position behavior of neural nets (not augmented with any search) as well as more recently what happens when you do put an "mcts" layer in front, I find it interesting to see what the neural nets are good and bad at.

Maybe this would be different for the very top available nets as I haven't tested them (e.g. ELF and the very latest LZ nets), but as far as I can tell, even moderately large neural nets I've trained on pro games or on Leela Zero data are noticeably worse than me (AGA 3d-ish) on average at snap-recognition of local tactics and tesuji and life-and-death-related moves. A wide variety of local fights that I solve "on sight" the policy net is noticeably uncertain about. The value net raw output is also not always good at recognizing the status of corner groups or capturing races, and sometimes requires the position to be played out a bit more before it becomes confident of the situation.

The raw policy net (still with no search) is better than me at having the instinct for the best-shape move in open-space situations. And in things like "do I keep extending or hane now?", or "do I jump or turn here" where it comes down to your feeling for the possible results and reading might only be of limited help if your intuition isn't already nudging you towards the answer. And the raw value net is massively better than me in judging large-scale influence-territory tradeoffs, strength of groups in running fights, and and all the other major parts of evaluating the board in the early midgame.

I find it actually sort of interesting that my personal "policy net" (and probably that of any other mid-dan player) with no conscious reading can be more accurate in sharp tactics and life and death, or at least be competitive. Unsurprisingly, when you add in search though, the whole thing becomes incredibly strong, as the search solves precisely those tactical things that the neural net was worse at, and then the power of the neural net for open-space and overall global judgment absolutely shines.

Maybe it would be different for a pro, but I've definitely learned some things about whole-board judgment from all this as well (and maybe have improved by about a stone at a result). The way I look at and value different kinds of shapes in the early midgame has changed quite a bit in a way that's hard to put into words, and I attribute it mostly to osmosis from just seeing example after example over time of what these bots say about different positions.

For me at least, what terminology to use or figuring out how to conceptualize something and verbalize it hasn't been relevant at all. Probably learning could be faster if we could wave a magic wand and have Go bots understand human language and how to "teach" in English or Japanese or whatever, but that's certainly not a requirement. Simply interacting with bots a lot has already let me improve my own intuition a little in the facet of the game where the bots seem to be the best, and interestingly that facet is *not* one of reading or calculation.

Why do you think it has come to mean that? That a direction isn't a vector is part of the meaning of the words, since a vector is direction plus magnitude. So I don't know that I'd think your choice of term was at fault.

I took a brief look, btw: and I found that Charles Matthews glossed Kajiwara as thinking of direction as a vector: https://senseis.xmp.net/?DirectionOfPlay%2FDiscussion. So perhaps this is a common misconception. But it's a weird one.

_________________
Occupy Babel!

Remembering always that we are talking about very subtle (and putative) differences between possible moves - as part of the effort to grasp what AI might teach us - I think talking about direction of play as meaning development or movement or haengma or the like is on the wrong track. Pinpointing the move that follows an assessment of the direction of play (as in western current usage) is a way of adding or over-emphasising magnitude. Or, better, if we try to follow O's thinking as he doesn't mention direction of play, it is making us think about the locale of the next move. It tilts us towards the local as opposed to the whole board.

I repeat what O has said. This method of using size of moves as a measure has been incredibly useful for us. Even where AI prefers a different move it does not automatically mean that the human move is bad. The AI move might only be microscopically better. But it is precisely that smidgeon of improvement that we are trying to understand. It seems to me (and I think this is the thrust of what O is saying) is that we may need to take several large steps back in order to move that inch forward.

That, by my reckoning, includes re-assessment of direction of play. What I believe has happened is that we players (possibly pros, too) have become accustomed to treating the direction of play as a way of finding the next move (i.e. emphasising development or movement), but what that entails is two questionable things. One is viewing a stone or group of stones as having a dominant direction (which is partly why it becomes a vector), instead of remembering that that stone/group exudes power in every direction. (I have already said that its implied plurality in the Japanese original is one aspect that is being lost.) I think even Kajiwara is guilty of this. The other glitch comes about because of thinking of playing from that stone/group. We don't do this when we have thickness - we play to drive the enemy towards it. But I mentioned in another thread a recent book which tried to emphasise that even a single stone has thickness, and I think that is what is being forgotten by us but picked up (invisibly almost) by AI programs. And, as I say, we may need to change our mindset radically to find that extra inch.

To try to wrench thoughts away from the go board and all the fixed but different associations we all have with it (e.g.. development may mean different things to different people), let me transport you to a different world.

There is a good Scots word 'airt' which means direction or quarter. Rabbie Burns uses it in the following poem. (blaw = blow, bonie = bonny and lo'e [pronounced loo] = love).

Of a' the airts the wind can blaw,
I dearly like the west,
For there the bonie lassie lives,
The lassie I lo'e best:

Following the usual stereotype you might think this is about a lovelorn poet who is about to embark on a journey to the west towards his lady love. This would be an analogue of the stereotypical thinking about developing from a stone/group on the go board.

In reality Burns had moved from a barren farm to another part of the country and was tilling and renovating a new farm to bring his wife to. So the true "direction of play" here was the reverse of the stereotype. As with thickness, this has its analogue in go, apparently overlooked in the case of single stones or small groups.

Now a different usage, but still by a lovelorn Burns.

O wert thou in the cauld blast,
On yonder lea, on yonder lea,
My plaidie to the angry airt,
I'd shelter thee, I'd shelter thee;
Or did Misfortune's bitter storms
Around thee blaw, around thee blaw,
Thy bield should be my bosom,
To share it a', to share it a'.

Here he talking about using a strong position not to develop but to defend - he is saying he wants to use his plaid (tartan cloak) to shelter his love from the direction the cold and angry wind is blowing (for the rest, bield = shelter and you need to be well up on your subjunctives). I suggest this too has often overlooked analogues in go.

Let us now return to the go board and consider this position from O Meien.



I was very much taken aback to find that this very ordinary looking position has never appeared in pro play. I triple checked in amazement. Although O said nothing about why he chose it, I assume it was because he wanted a tabula rasa to make his point. Which was that a move around A is the duh move here. He was talking only about size of move, on the basis of the sort of thinking that goes "if I don't play there he will play there and that's HUGE for him." I'm sure we all recognise that approach. If we add direction of play to the mix (which he did not) and apply it in the usual way, the case for A becomes overwhelming. After all the direction of play for each Black corner group is down the right side and to let White scupper two directions at once must be suicidal, no?

O, trying to approach this from an AI perspective, suggests a different way of playing, as follows. He plays keima in the top left, allows the double extension for Black on the right, then does an AI shoulder hit at A.



He says the shoulder hit this early would not have previously occurred to pros because it allows Black B or C, both of which confirm Black's territory (and are, of course, ajikeshi). (But, trying to think like an AI, he goes on to postulate Black E now instead of B or C.)

Now what I get when I run LeelaZero is that the AI essentially agrees with his new way of thinking. The difference is that it chooses the keima at D (and not c14) as its best move, and even at this stage it rates the shoulder hit as second best. The wariuchi on the right is certainly considered but is microscopically lower (52.1% against 52.4% but these tiny edges are what we presumably need to focus on).

It might shock O but LZ barely considered his White keima, so I would infer he needs to think even more about some aspect of play in that corner. My guess is that the LZ keima protects White better from the angry airt of the directional power of Black's upper right shimari. O may also be shocked to learn that after this keima, LZ does not opt for the double extension the right side. It prefers E, though in this case the difference between E and a move in the right centre is nanoscopic rather than microscopic.

LZ, after its White keima at D, looks at a clutch of moves in the right centre and also a clutch of moves (though rated lower) in the upper left (including another early shoulder hit), but E is a singleton in that area. Possibly it sees singleton moves as more urgent, or prefers the certainty of a single choice that maybe allows a deeper search (???).

Anyway, I think direction of play (among other things) merits some re-appraisal. But, if you allow me to share an impression I often get, I can't shake the idea that, as here, suggestions for new thinking are resisted on L19 in a knee-jerk way, and I find that strange. Even if you think about something and come to the same conclusion you had before, the very act of thinking has made you stronger. And that leads to a another new idea: direction of thinking...

Now I cannot help but think of each stone as having reverse gravity, or negative gravity, like a white hole. Putting stones together increases their negative gravity (decreases their gravity). If you have groups on the board, they repulse each other so that new stones would rather be on open areas.

Here's a demonstration of a white hole at 7:00


(it is likely erroneous but logical to think of big points as deep and thick stones or groups as mountainous in this fashion).

_________________
On Go proverbs:
"A fine Gotation is a diamond in the hand of a dan of wit and a pebble in the hand of a kyu" —Joseph Raux misquoted.

Maybe two main factors contribute to competence in most single-player competitive fields of endeavour including active sports and mindsports: field-specific skill, in which humans excel pound to pound, and interdisciplinary skill, mastered by machines not just in calculation ability but emotional control Machines train their field-specific skill many times more efficiently than any human can until it surpasses any human's field-specific skill.

_________________
On Go proverbs:
"A fine Gotation is a diamond in the hand of a dan of wit and a pebble in the hand of a kyu" —Joseph Raux misquoted.

Well, I just mentioned Ishida in passing, maybe unfairly.



Well, he does get into middle game plays and endgame plays, as well.

It is important to note a couple of things. First, Ishida is using deiri values, so to find how much a play gains you have to divide by 2. Second, he is offering a refinement of traditional values. For instance he starts off by saying that the first move in an empty corner is worth 20 pts. That is consistent with the traditional value of a handicap stone as 10 pts. (20/2 = 10. )



I don't know when the book was first published. But I don't think that it was until the late 1970s that people began to realize that the traditional values were too small. I estimated the value of a handicap stone at around 14 in 1976, but I didn't publish anything. In 1977 the AGA Journal published an article based upon statistics claiming that correct komi is 7 pts., which is consistent with a value of 14 pts. for a handicap stone. That also indicates that the first move in an empty corner is worth 28 pts. in deiri values. So Ishida was on the right track, he just didn't go far enough.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.