Human vs AI in Go in 2026: Lessons for superintelligence?

In July 2026, Korean Go player Shin Jin-seo defeated KataGo 2-1 , with only [1] a 2-stone handicap (favoring Shin). I thought this was very surprising, and posted on LessWrong and Twitter asking how to interpret this. I didn’t get any entirely satisfactory answer, so I decided to spend ~one day looking into it and writing up my understanding. To preface this post: I don’t know all that much about Go and I'm a quite bad Go player (like, no better than 20 kyu or something like that). But in 2022 I did an interpretability project on a Go-playing model and have spent quite some time thinking about Go-playing AIs. Why I found this surprising Go is the canonical domain where AI models are considered to be vastly, vastly superhuman. We heard it from the best Go players themselves, who have faced off and been defeated by these Go-playing AIs. Ke Jie, the professional Go player then ranked world No. 1, said in 2017 after losing to AlphaGo that “AlphaGo to me is 100 percent perfection, to me AlphaGo is the god of the Go game”. I've been interested in Go-playing AIs partly because it serves as an intuition pump for me for superintelligence in other domains: Just imagine if an AI model became