It is currently Sun Jan 26, 2020 12:57 am

All times are UTC - 8 hours [ DST ]




Post new topic Reply to topic  [ 1 post ] 
Author Message
Offline
 Post subject: The value of particular Go books in the AI age
Post #1 Posted: Thu Dec 12, 2019 9:44 am 
Tengen

Posts: 4875
Liked others: 0
Was liked: 661
Introduction:

This thread is related to the thread "The value of Go books in the AI age" at https://lifein19x19.com/viewtopic.php?f=15&t=17127 but here, in the go books forum, discussion studies PARTICULAR go books inhowfar they might need to be reevaluated in the AI age.

I start with some of my own go books. I have browsed their theory to answer the study question. Examples represent truths if they apply theory comprising truths; such examples do not need reevaluation. I have not studied the other examples by AI yet, also because I do not have local hardware suitable for AI yet. Many examples are local, so AI analysis may be difficult.

For now, I have browsed my two Positional Judgement books - Vol. 1 Territory and Vol. 2 Dynamics - and four Endgame books - Endgame Vol. 1 Fundamentals, Endgame Vol. 2 Values, Endgame Vol. 3 Accurate Local Evaluation and Endgame Problems Vol. 1. Positional Judgement 1 + 2 and Endgame 1 appeared before AlphaGo. Endgame 2 + 3 and Endgame Problems 1 appeared after the advent of AlphaGo but were not influenced by the existence of AI at all.


All Six Books:

The theory remains correct in the AI age.


Positional Judgement 1 + 2:

Not only does the theory remain valid - but the books also anticipated some AI strategies (I have explained that before) and some scepticism of what other books might need to consider (in particular several principles in the theory for invasions in potential territory regions).


Endgame 1 Fundamentals:

This book for kyus has simplifying theory. Dan players need to also learn exceptions beyond the scope of the book when application of theory becomes more advanced. Concerning the book's theory, AI just shares the demand of dan players.


Endgame 2 + 3 + Endgame Problems 1:

Most (Endgame Problems 1: all) of the theory relies on mathematical truths. The theory applied before the microendgame has the exceptional approximation of ignoring the global consideration of the one last point of the microendgame. Ko exchanges use another approximation when comparing move values. A small part of the theory peeks at strategic concepts of the middle game so amounts to principles rather than truths. AI (and professionals and amateurs) should learn from the books' theory representing truths - not vice versa.


This post by RobertJasiek was liked by: Marcel Grünauer
Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 1 post ] 

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