https://lifein19x19.com/viewtopic.php?p=276494#p276494 to

The point is a higher meta-level sente always takes precedence over a lower level sente. The easiest demonstration for this that amateurs would be familiar with would be in the endgame. A kyu player might see multiple sente moves on the board and think that means they're all miai. But if you play a sente move where the opponents response is sente, then that sente move had already lost the meta-sente. Even if the kyu player maintains sente in the endgame they might lose points by choosing an order of moves that loses the meta-sente. If language wasn't so important, no one would read books. Even before AlphaGo people would have been getting stronger only through playing games!

As someone who has been interested in how we internally interpret language recently since my literary ability is not what it used to be when I was a teenager, I'd be interested in an experiment where we get 64 beginners from a wide range of backgrounds to play 8 games each every day on Fox, Tygem, Oro, Wbaduk, Yuugen no Ma, Pandanet, KGS, OGS, so 64 19x19 games a day of intense training, for 64 days, and after each game they must annotate their thoughts. They're not allowed to read anything to do with Go to avoid any possible linguistic bias in their thinking. It would be very interesting to see how they play, how good they get and the language they use when reviewing each game.

And we can do the same for Table Tennis and Piano. I deliberately want to avoid learning music theory to see if I can play the 2nd ending of Hikaru no Go by ear. Something similar would be getting Table Tennis and Tennis pros to play each others games, Or even the Xjianshochess I made, haha!

Churn!

Churns are useful for storing milk

_________________
North Lecale

The Churney continues! The Churney to becoming an Igo master!

A rich palette of Go vocabulary might be more useful for appreciation and description of Go games rather than actually playing.

An analogy may be that you can tap into a rich vocabulary for describing an oil painting in terms of its composition, how an artist sits in historical context, an almost infinite range of terms for colours and tones, and technical terms for chemical composition of the paint and canvas used. With technological support, this vocabulary can even extend to describing what plans the artist originally had for the work (revealed by x-ray), or with the help of AI, what other versions may have looked like if they had been taken forward,

However, none of this might help you become a more proficient painter - and leading artists themselves may be quite uninterested in such commentaries and see them as 'niche'.

I am often both amused and bemused by the fair number of posts on L19 that make arguments based on the supposition that its readers are striving to be pros.

99.9% of go players are and always will be amateurs, with relatively little time. Compared with those who are pros or who are talented or young or lucky enough to embark on the path to prodom, they need different ways to improve either their strength or their appreciation of the game. The pros themselves decided words were the way to do it. Over 500 years ago, the Chinese produced "model" books and invented new terms. In more modern times, Honinbo Shuho made the Hoensha magazine (i.e. words) the linch-pin in his efforts to spread the game nationally in Japan. A few decades later, the Nihon Ki-in went a step further in the 1920s and got professional journalists to enlarge the go vocabulary. In recent times, more effort seems to be going into videos than paper, but words are still the medium of communication. Even pros who became strong without reading the James Davies book on tesuji or the like spoke to teachers and co-students, and can, and do, all speak fluent Go.

You can have a debate about how and how much to use words in learning or teaching go, or which language to prefer, but to deny words their central place in go is like having fish and chips without the fish (or the chips!).

I was watching Choi Jeong's game and I think this is probably one of the best demonstrations of what you mean, since throughout the game AI had her ahead by two points but with a very high percentage of winning! Which is typical AI play—She had complete control of the game, initiative, even if the lead in points was small it was never going to go away!

What do you mean by "higher meta-level sente" and "lower level sente"? One can distinguish local sente (in a local region) and global sente (a local sente currently is one of the equally best moves on the board) but is this what you mean? One can also distinguish the initiative, which refers to playing successive sente moves on different parts of the board, from sente, which refers to a region.



If the opponent's local reply is his sente, then it was not the player's local sente. If the opponent's reply elsewhere on the board is his global sente, then the player's local play might have been a local sente but currently not a global sente.



Don't you just mean the "initiative"?



Does he, or does he not, maintain sente aka the initiative aka, your superfluous phrase, meta-sente?

Yes, since amateurs tend to mistakenly use sente when referring to initiative since they're etymologically similar, I thought of that term.

I'm pretty sure there are special situations were one player can play a move the other must respond to, but then that response also must be responded too.

I would argue that Neural Nets are nothing more than very large dictionaries of advanced and precise words.