New pro teaching in the AI age?

[John Fairbairn]

[Translation - Part 3]

SuzukiDiag6.GIF (45.53 KiB) Viewed 11211 times

Diagram 6 (Wedge)
Rather than White 2 in Diagram 5, I would want to try the wedge-in at White 1. If Black responds at 2 and 4, White forces at 5 and so has made something of a gain. After Black 6, White plays 7 and has a secure shape.

SuzukiDiag7.GIF (46.08 KiB) Viewed 11211 times

Diagram 7 (∆ is at a good point)
If Black resists White 1 with 2 and 4, White 5 to 9 is a fixed pattern. In this case the ∆ stone is on a good point and so 'A' and 'B' next are miai for him.

SuzukiDiag8.GIF (47.08 KiB) Viewed 11211 times

Diagram 8 (White has been strengthened)
Continuing, Black 10 is a tesuji and so he can at least play on two sides, at 12 and 14, but after White 19 White has become extremely strong, and it looks as if he can aim at the invasion at 'A'.

[John Fairbairn]

[Translation - Part 4]

SuzukiDiag9.GIF (44.91 KiB) Viewed 11210 times

Diagram 9 (A one-space extension is too narrow)
After White 1 and Black 2, unlike the open-skirt approach of Diagram 4, White's extension to 3 is probably a little narrow. Black can attack at 4 and 6 and in the process be building up a moyo on the lower side.

SuzukiDiag10.GIF (44.24 KiB) Viewed 11210 times

Diagram 10 (Approaching one step further)
In this position, it is also interesting to approach one step closer down with White 1 [C of the Topic Figure]/ If Black 'A@, White extends to 'B' and has even more room than in Diagram 3.

[The next sentence in the original seems to have been misplaced. I have moved it to * in the next part.]

SuzukiDiag11.GIF (45.08 KiB) Viewed 11210 times

Diagram 11 (Attachment and pull-back)
Black blocks downwards with 2. After the attachment and pull-back of White 3 and 5, White can settle himself with the slide to 7. Next White 'A' at the 3-3 point or the jump to 'B' are miai. will be happy with the sequence of this diagram.

[John Fairbairn]

[Translation - Part 5]

SuzukiDiag12.GIF (45.64 KiB) Viewed 11208 times

Diagram 12 (Black's counter-attack)
However, there is a risk that Black may answer White's attachment 1 with the counter-attack 2 and 4. White must be sure to answer these moves reliably.

SuzukiDiag13.GIF (46.52 KiB) Viewed 11208 times

Diagram 13 (Sabaki achieved)
Continuing from Diagram 12, if White can settle the situation with 7 and 9 and then attach at 11, he will have coped well [made sabaki] overall on the right side. If he then more or less copes in the corner, his four stones including ∆ can be treated as light [i.e. can be sacrificed] and will not be an issue. As one example, if Black later hanes at 'A', White cuts at 'B'. Then if Black 'C', White 'D', Black 'E', White now has 'F', Black G', White 'H'. Although this diagram may be rather difficult, White would be happier with this rather than simply with the sequences of Diagrams 3 or 4. [*] The later sabaki (coping), as in Diagrams 12 and 13, may be a little difficult, but among pros many would would be likely to play this way.

[Gérard TAILLE]

I'm not investing in the exercise as outlined but I'm offering shodan thinking, or the way mine goes in this case. I'm highlighting a couple of non-Japanese terms.

$$Wc
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . 5 . |
$$ | . . . . . . . . . . 0 9 . . O O O 4 7 |
$$ | . . . . . . . . . . . 8 O O X O X 2 . |
$$ | . . . O . . . . . , . . X X X X 1 3 . |
$$ | . . . . . . . . . . . . . . . . 6 . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |

Click Here To Show Diagram Code

[go]$$Wc
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . 5 . |
$$ | . . . . . . . . . . 0 9 . . O O O 4 7 |
$$ | . . . . . . . . . . . 8 O O X O X 2 . |
$$ | . . . O . . . . . , . . X X X X 1 3 . |
$$ | . . . . . . . . . . . . . . . . 6 . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |[/go]

I see that the exchange is good for black but what about the following answer:

$$Wc
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . O O O 4 . |
$$ | . . . . . . . . . . . . O O X O X 2 . |
$$ | . . . O . . . . . , . . X X X X 1 3 . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X 5 . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |

Click Here To Show Diagram Code

[go]$$Wc
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . O O O 4 . |
$$ | . . . . . . . . . . . . O O X O X 2 . |
$$ | . . . O . . . . . , . . X X X X 1 3 . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X 5 . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |[/go]

[CDavis7M]

The later sabaki (coping), as in Diagrams 12 and 13, may be a little difficult, but among pros many would would be likely to play this way.

You know, AI made the opening relatively easy. There are fewer joseki played and the joseki are shorter and simpler. So that was nice. But now we see AI increasing the difficulty with Suzuki suggesting point C (at least that's what Pros would play), even if it looks a little difficult. Diagrams 12 and 13 from C do look a little difficult! B seems to lead to an easier/safer sequence and is probably what I will remember. But that comes back to the point of your post: in the past a pro might have recommended something simple but now they are presenting a more nuanced assessment.

I hadn't thought about it before but that seems to be the case here. So it's not just that the Lecturer is presenting an AI suggestion given the prevalence of AI (which happens), but also that the Lecturer is providing a more complex human assessment of the position compared to pre-AI teaching.

[jlt]

It's not so much whether A is better than C or vice-versa which is interesting here, but the analysis of the position and the implications of each choice.

Here is a variation discussed by Suzuki until , Suzuki considers it as bad for White but here is a continuation that Katago likes for White:

$$Wc
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . O O O . . |
$$ | . . . . . . . . . . . . O O X O X . . |
$$ | . . . O . . . . . , . . X X X X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . 3 . . |
$$ | . . . . . . . . . . . . . . 4 . . 0 . |
$$ | . . . , . . . . . , . . . . . 6 8 9 . |
$$ | . . . . . . . . . . . . . . 5 7 1 . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . 2 . . |
$$ | . . O . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . , . . . . . X . . . |
$$ | . . . . . X . . . . . . . X . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------

Click Here To Show Diagram Code

[go]$$Wc
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . O O O . . |
$$ | . . . . . . . . . . . . O O X O X . . |
$$ | . . . O . . . . . , . . X X X X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . 3 . . |
$$ | . . . . . . . . . . . . . . 4 . . 0 . |
$$ | . . . , . . . . . , . . . . . 6 8 9 . |
$$ | . . . . . . . . . . . . . . 5 7 1 . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . 2 . . |
$$ | . . O . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . , . . . . . X . . . |
$$ | . . . . . X . . . . . . . X . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

$$Wcm11
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . O O O . . |
$$ | . . . . . . . . . . . . O O X O X . . |
$$ | . . . O . . . . . , . . X X X X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . O 2 . |
$$ | . . . . . . . . . . . . . . X . . X . |
$$ | . . . , . . . . . , . . . . . X X O . |
$$ | . . . . . . . . . . . . . . O O O 1 . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X 9 . |
$$ | . . O . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . 8 3 7 . |
$$ | . . . O . . . . . , . . . . . X 4 5 . |
$$ | . . . . . X . . . . . . . X . . . 6 . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------

Click Here To Show Diagram Code

[go]$$Wcm11
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . O O O . . |
$$ | . . . . . . . . . . . . O O X O X . . |
$$ | . . . O . . . . . , . . X X X X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . O 2 . |
$$ | . . . . . . . . . . . . . . X . . X . |
$$ | . . . , . . . . . , . . . . . X X O . |
$$ | . . . . . . . . . . . . . . O O O 1 . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X 9 . |
$$ | . . O . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . 8 3 7 . |
$$ | . . . O . . . . . , . . . . . X 4 5 . |
$$ | . . . . . X . . . . . . . X . . . 6 . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

[John Fairbairn]

jlt: A few disparate points, on either side of the fence.

1. Since we keep on hearing tales of bots changing their minds and agreeing more with humans if we increase the number of playouts (or whatever the current word is), I suppose we have to ask whether that's a factor here.

2. Assuming it's not, I think we need to note that Suzuki did not actually say good or bad - he said Black can cut through immediately. The point there, as I saw it, was that it's a very simple, forcing line of play of the type that humans can follow and be happy about. Pragmatism, in other words. Play only what you can understand,

3. But most interesting to me was that I looked at the final position and my first response was: How on earth can that be good for White? Low on the second line, and all that. But we know that bots don't seem to mind that as much as we do, and so I thought a little more. Then it dawned on me that the White group in the lower right is actually a Go Seigen group. That means we can write off the whole of the "contaminated" lower-right quadrant in territory terms. If we then look at the rest of the board, Black has more than White in the upper right, but on the left side White has a 3:1 majority. So, the idea that White might stand better here suddenly didn't look so fanciful after all! Could Suzuki have a blind spot about GSG groups? They do seem characteristic of only the very best players - Go, Cho Hun-hyeon, Honinbo Shuei - and bots, of course.