# Google DeepMind 成立 DeepMind Institute，跨学科研究 AGI 安全与治理

- 来源：The Decoder：AI News（RSS）
- 作者：Matthias Bastian
- 发布时间：2026-09-17 01:00
- AIHOT 分数：45
- AIHOT 链接：https://aihot.news/items/cmu4d9gsd0jb5ro4wx4tk4z27
- 原文链接：https://the-decoder.com/google-deepmind-launches-interdisciplinary-institute-to-tackle-the-big-questions-around-agi

## AI 摘要

Google DeepMind 成立 DeepMind Institute（DMI），联合 Google 及全球科学界，从艺术、人文与政策等跨学科视角研究 AGI 的安全、治理及网络攻击、失控等风险。

## 正文

Google Deepmind has founded the Deepmind Institute (DMI), a platform for interdisciplinary research and debate around artificial general intelligence (AGI). The institute brings together researchers from Google Deepmind, Google, and the global scientific community to tackle questions about safety, governance, and risks like cyberattacks or loss of control. The answers shouldn't come from technologists alone but also from the arts, humanities, and policy, according to directors Shane Legg, James Manyika, and Nobel laureate Demis Hassabis.

There's no agreed-upon definition of AGI. Deepmind describes it as a system with all the cognitive abilities of the human brain and says AGI is close, even though today's AI systems still fail at simple tasks and lack creativity. Hassabis sees it coming within a few years, Legg recently said a precursor could arrive by 2028. OpenAI focuses on economic value, and CEO Altman believes AGI could come by year's end. AGI shouldn't be confused with "ASI", an artificial superintelligence that would far surpass humans in every respect, though AGI is likely a necessary step toward it.

Deepmind Institute