Human diffusion · 21
How long does an AI joseki take to become a human habit?
AI can raise novelty in professional decisions while concentrating repeated opening patterns. The findings coexist because one measures a move's newness and the other measures whether a population repeats the same opening repertoire.
Define adoption before looking for a first game
One appearance may be an experiment. Repetition, spread across players, textbook entry, and later decline are different-strength signals and belong on separate points of the curve.
Board A is one trial; board B marks the same structural choice repeating across players. More novelty and less diversity can coexist
One PNAS study of more than 5.8 million professional decisions finds increased post-AI novelty associated with better decisions. Another, using about 70,000 games and 15 million moves, finds advancing knowledge alongside concentrated common patterns.
The move-level novelty axis can rise while population opening breadth contracts. Early 3-3 shows surprise compressing into a default
DeepMind identified early 3-3 as an AlphaGo innovation in 2017 and noted rapid professional trial. Adoption meant more than copying coordinates: players accepted fewer exchanges and unfinished aji as a timing choice.
The older line settles locally; the newer line buys whole-board initiative with fewer moves. Elite players usually move before the textbook
Professionals have stronger analysis, lower information cost, and denser observation of peers. A diffusion study should separate professionals, strong amateurs, and ordinary amateurs instead of reporting one aggregate frequency.
Three population timelines expose a measurable lag for the same pattern. Coordinates spread faster than reasons
Players may copy the opening moves while retaining the old purpose. The concept has diffused only when tenuki timing, ladder conditions, and return points change with the coordinates.
Two continuations after the same first six moves reveal whether the old habit still governs. Platforms, rules, and komi price experimentation
Public records, streams, and analysis software reduce observation delay. Server rules, rank mix, and time controls change the cost of trying a novelty, so adoption rates need environmental metadata.
The same local idea carries a different trial cost in blitz and long games. Build a rerunnable joseki diffusion ledger
For each normalized local pattern, store first-seen interval, repeat players, group-specific adoption, decline point, and mirrored coordinates. That separates a fashion spike from a durable habit.
Every step from one board event to a population S-curve remains auditable.
Research and primary sources
When studying joseki diffusion
- Separate first appearance from mainstream adoption
- Separate move novelty from population diversity
- Stratify professional and amateur play
- Track coordinates and rationale
- Keep rules and platform metadata