The article also highlights that the shift in performance happened 18 months after the introduction of AlphaGo, coinciding with the release of Leela Zero, an open-source Go engine. This allowed players to build tools that showed the AI's reasoning when picking moves and enabled massive input learning. The author concludes by suggesting that AI could help humans push through their limits in more domains, as seen in the game of Go and chess after DeepBlue's victory over world champion Kasparov in 1997.
Key takeaways:
- Professional Go players have significantly improved their skills after the introduction of AlphaGo, an AI that could beat the best human Go players.
- About 40 percent of the improvement came from moves that could have been memorized by studying the AI, but moves that deviated from what the AI would do also improved, accounting for 60 percent of the improvement.
- The trend shift in Go happened 18 months after AlphaGo, coinciding with the release of Leela Zero, an open source Go engine that allowed players to build tools and do massive input learning.
- Similar to the chess world after DeepBlue beat the world champion, the introduction of AI in Go did not stifle human creativity but instead led to a flourishing of skills and creativity among human players.