跳到主要内容Skip to main content

通用人工智能快到了AGI Is Just Around the Corner

微型指南 · 约 5 分钟Guide · about 5 min · 引用核验 2026-11-03(仅引用链接与日期) · refs checked 2026-11-03 (reference links and dates only)

「临近」是说法,不是可核对的里程碑。"Soon" is a claim, not a verifiable milestone.

各家都在说 AGI 临近,说明它有一个可核对的到达时间。Everyone says AGI is near, so there must be a checkable date when it arrives.

AGI 的定义与评价标准并未统一。“临近”只有在能力范围、测试条件和完成标准明确时才可核对;研究目标或时间预测不能直接当作已经实现的证据。AGI definitions and evaluation criteria are not settled. A claim of proximity needs specified capabilities, test conditions and completion criteria; a research goal or forecast does not establish achievement.

例子与判断Example and interpretation

以下是教学情境,不是本站运行或实测记录。A teaching scenario, not a measured or executed site test.

看到“明年达到 AGI”时,先问:AGI 在这里指哪些能力,怎么测,谁复核,日期是承诺还是预测?回答不了这些问题,就把它记作预测,不能当作项目排期的依据。For “AGI next year,” ask which capabilities AGI means, how they are measured, who checks them, and whether the date is a commitment or prediction. Without those answers, record a prediction rather than a planning assumption.

出处Sources
  • Morris 等《Levels of AGI for Operationalizing Progress on the Path to AGI》(Google DeepMind,ICML 2024)Morris et al., Levels of AGI (Google DeepMind, ICML 2024) ↗
  • Turing《Computing Machinery and Intelligence》(Mind, 1950,Oxford Academic 官方页)Turing, Computing Machinery and Intelligence (Mind, 1950) ↗
围绕这些概念继续读More on these concepts

—