Why Taiwan 16-Tile Mahjong Still Has No AlphaGo Moment

by Nat Currier 8 min read
AIReinforcement LearningGame TheorySystems Design
Excerpt

If AI can dominate Go and chess, why has Taiwan-style 16-tile mahjong not produced a public, superhuman breakthrough model? The answer is not one missing algorithm. It is an ecosystem problem with technical, data, and incentive bottlenecks.

Cite This

Nat Currier. "Why Taiwan 16-Tile Mahjong Still Has No AlphaGo Moment." nat.io, 2026-02-17. https://nat.io/blog/taiwan-16-tile-mahjong-why-no-superhuman-ai

Taiwan 16-tile mahjong lacks a superhuman AI like AlphaGo due to an ecosystem deficit: unstable rules, poor data quality, missing standardized evaluation, and complex gameplay that complicates modeling and training.

https://nat.io/blog/taiwan-16-tile-mahjong-why-no-superhuman-ai

Share link (tracked): https://nat.io/blog/taiwan-16-tile-mahjong-why-no-superhuman-ai?utm_source=citation&utm_medium=referral&utm_campaign=blog_cite

Key stat: 8 minute read