Hassabis shortens AGI timeline to 2–3 years and proposes pre-release review
What happened
The Decoder reported on October 22 that Google DeepMind founder Demis Hassabis had earlier this year expected AGI within 3 to 5 years, and has now shortened that to 2 to 3 years. He defines AGI as a system matching the full cognitive range of the human brain, and proposes an "Einstein test": train a system only on knowledge from before 1901 and check whether it can reproduce Einstein's 1905 paper. On regulation, he proposes an industry-funded, government-controlled oversight committee to review models before release, and warns that bad regulation is worse than none.
Written by AI from the coverage · updated 2 hours ago
Coverage
Follow the reports to see the story from different sides.
- The DecoderDeepmind founder Hassabis shortens AGI timeline from five to two to three years
Google DeepMind 创始人 Demis Hassabis 将 AGI 到来的预期从今年早些时候的3至5年缩短至2至3年。其定义是系统匹配人脑的完整认知范围，并提出“爱因斯坦测试”，即仅用1901年前的知识训练的系统能否重现爱因斯坦1905年的论文。他近期还提议设立由行业出资、政府控制的监督委员会，在模型发布前进行审查，并警告糟糕的监管比没有监管更糟。
Heat over time
Not enough continuous observations to draw a trend yet.