Google傘下のDeepMindが極秘に進めた「マリオ計画」とは何だったのか、AGI安全策の行き詰まりが明らかに - GIGAZINE
https://gigazine.net/news/20260403-google-deepmind-project-mario-agi-safety/The evidence pack includes HTML, screenshots, summaries, and metadata. It can be downloaded on Pro.
Google傘下のDeepMindが極秘に進めた「マリオ計画」とは何だったのか、AGI安全策の行き詰まりが明らかに - GIGAZINE
Open the archived HTML with saved-time metadata attached.
This HTML has CSS and images embedded, so it can still be opened even if the original page disappears.
This page reports on Project Mario, a secret initiative DeepMind launched from autumn 2015 to establish governance structures preventing AGI from being controlled solely by parent company Google. Co-founders Demis Hassabis and Mustafa Suleyman explored multiple safety measures, including independent boards and external funding options. However, Google opposed these proposals, particularly after CEO Sundar Pichai emphasized AI's importance to core businesses. Despite continued negotiations through 2018, DeepMind failed to achieve the independence it sought, ultimately relying on internal influence instead.
