Post-AI Leaps in Human Go Decision Quality, Novel-Move Rates and Strategic Entropy in the Thirteen-Chapter Classic of Go: chapter answers
After AlphaGo, do people grow duller by leaning on AI, and does Go become uniform? This book starts from the 2023 study by Shin, Kim, van Opheusden and Griffiths in the Proceedings of the National Academy of Sciences, which used KataGo to score 5,857,513 decisions by professional players from 1950 to 2021 and found that both decision quality and novelty rose clearly after AI appeared. The book sets out the paper's numbers as published and then asks, item by item, under what conditions they hold. Apart from the paper's figures, all parameters are illustrative and all calculations are generated…
DOI pending · monograph page goes live after publication · 5DL-B-230 · 中文版
- Chapter 1 · The Leap at the Break: a Decision Quality Index on a KataGo Yardstick
Chapter 1 turns the paper's Table 1 into effect sizes: after AI an ordinary move's Decision Quality Index is 0.59754 points higher, a novel move's 1.11258 points, and the old penalty for novelty shrinks by 0.848; it als… - Chapter 2 · A Break in Time: Regression Discontinuity at March 2016
Chapter 2 demonstrates regression discontinuity in time on simulated data: with a true jump of 0.30 a 12-month bandwidth detects it only 0.685 of the time, while a world with nothing but a ramp is declared significant w… - Chapter 3 · Novel-Move Rates: the First Historically New Move, a Growing Database and the Memorisation Test
Chapter 3 shows that the Novelty Index drifts mechanically as the historical database grows and tests the inserted-AI-games check against memorisation: copying particular games is fully exposed, but learning the AI's op… - Chapter 4 · Strategic Entropy: Joseki Concentration, Local Optima and Conformist Lock-In
Chapter 4 measures joseki concentration with Shannon entropy; the strategic entropy of an illustrative corner position rises from 1.608 to 2.360 bits, and a model of conformist learning shows an outside evaluator bringi… - Chapter 5 · The Contest over Mechanism: Internalisation, Memorisation and Composition
Chapter 5 separates internalisation, memorisation and composition, showing that player fixed effects push a spurious interaction from 0.831 down to -0.053 but cannot touch memorisation. - Chapter 6 · The Limits of the Yardstick: Evaluator Noise, Engine Versions and the Win-Rate Ceiling
Chapter 6 tests the yardstick: evaluator noise averages out in large samples, and the same 3-point mistake costs 14.566 points of win rate in an even position and 1.431 when far ahead. - Chapter 7 · Classics East and West: the Classic of Go in Thirteen Chapters and Franklin's The Morals of Chess
Chapter 7 weighs Zhang Ni's Classic of Go in Thirteen Chapters and Franklin's The Morals of Chess, computing that the break-even point of "rather lose several stones than lose the initiative" falls from 7.00 stones in t… - Chapter 8 · Closing the Loop: a Team Routine for AI Reviews and an Audit of Editorial Decision Quality
Chapter 8 gives seven rules for team reviews with AI, an audit of decision quality in the publisher's AI-assisted writing, the distilled Review Quality Scorer and a pre-registered protocol. Marked boxes separate publish…
Illustrative model calculation, not a live quote