Go Players Disempower Themselves to AI (www.lesswrong.com)

In a compelling exploration of the intersection of AI and the game of Go, the article reflects on the impact of AI tools like Leela on player behavior and tournament integrity. Following the landmark defeat of top Go player Lee Sedol by Google DeepMind’s AlphaGo in 2016, the cultural narrative around Go remained largely unchanged, with AI seen as a complement rather than a replacement for human players. However, as AI usage proliferated, particularly through platforms like Leela, concerns about cheating arose, culminating in a controversial disqualification of player Carlo Metta in 2018. While initially banned for alleged AI-assisted play, he was later exonerated, revealing a troubling trend of unregulated AI use and a community struggling with the implications of these new tools. The significance of this situation lies in its reflection of broader challenges within the AI/ML community, particularly regarding ethical considerations and the nature of human achievement. As many Go players began to rely on AI for competitive advantage, some rationalized this dependence as a means of artistic expression, leading to a disconnect between genuine skill and algorithmic support. The ease of cheating and the community's resistance to addressing it highlight an urgent need for clear frameworks to ensure fair play and preserve the integrity of Go as a competitive and cultural endeavor. The article underscores the importance of navigating these complexities thoughtfully as AI continues to shape the experience and landscape of traditional games.