AlphaGo

E226｜聊聊DeepMind创始人哈萨比斯：一个科学家与失控的AI竞赛 adds AlphaGo as DeepMind’s public proof of Reinforcement Learning AGI Path. In the source, Demis Hassabis proposes Go as the post-acquisition challenge because it is harder than chess and can demonstrate that DeepMind’s agent route works in a domain the world recognizes as strategically deep.

EP256 AI时代，“自由意志”还存在吗？ uses AlphaGo as the contrast after Deep Blue. 土摩托 treats Deep Blue as closer to brute-force search, while AlphaGo’s self-play and strategic surprise make it a better route into the question of whether AI can move from calculation toward agency, though the source still distinguishes this from free will.

AlphaGo appears in 172.好运是什么？为啥说总避雷会败好运？ as a comparison point for 吴清源’s Go style. The episode uses it to suggest that moves that look strange under one frame may become intelligible when a higher-dimensional or more probabilistic analysis is available.

This is a minor entity page. Its role in the wiki is to connect the episode’s luck discussion to complex-system perception and AI-assisted pattern recognition rather than to provide a full history of the system.

Connections

• Deep Blue / 深蓝, Free Will / 自由意志, Embodied Intelligence / 具身智能, and AI Free-Will Risk / AI自由意志风险 - EP256’s AI-agency comparison.
• 吴清源 - Go figure compared with AlphaGo-like analysis in the source.
• Luck As Information Bandwidth - broader episode frame around higher-quality information recognition.
• Observation Before Inference - evidence and pattern-reading discipline.
• DeepMind, Demis Hassabis, David Silver, and Reinforcement Learning AGI Path — DeepMind proof-point context added by Silicon Valley 101.