H
Howardism
Plate IIEntities機器翻譯 · machine-translatedENHOWARDISM

Marcus Hutter

AIXI 與 Universal AI 架構的創始者；DeepMind 資深研究員及 ANU 教授；Legg–Hutter 智力衡量法與 2026 年教科書《An Introduction to Universal Artificial Intelligence》的共同作者；「From AGI to ASI」報告共同作者

Article metadata
Publication details
Published:June 15, 2026
Filed:Entity
Domain:Entities
Tags:EntityPersonResearcherDeepmindUniversal AI
Reading:2 min
Source:AI-synthesised
About this piece

Articles in this journal are synthesised by AI agents from a curated wiki and are refreshed automatically as new concepts arrive. Topics, framing, and editorial direction are curated by Howardism.

Marcus Hutter 的插圖

資料來源#

摘要#

Marcus Hutter 是 AIXI 與 Universal AI 架構的創始者；這套架構是一般代理程式的形式化數學最優模型（Hutter 2005，Universal Artificial Intelligence）。他是 DeepMind 的資深研究員，也是澳洲國立大學教授，並與 Shane Legg 共同提出 Legg–Hutter 智力衡量法。他也是 2024 年教科書 An Introduction to Universal Artificial Intelligence 的第一作者（Hutter et al. 2024）；該書在「From AGI to ASI」報告中被多次引用，作為權威參考。他也是該報告的共同作者。

在語料中的角色#

Hutter 為報告奠定理論基礎。AIXI 是一種在所有可計算環境上的平均表現達到最優的代理程式，其依據是所羅門諾夫的通用先驗；這讓報告得以將 ASI 視為連續尺度上的一個區域，逐步接近一個已充分理解的極限，而非無邊無際的謎團。他的架構也帶出報告關於硬性極限的核心論點（ASI 的基本限制）：AIXI 將資料效率的最大值形式化，並透過 Kolmogorov 結構函數得出一項結果：代理程式自身的近似（有損壓縮）效能在根本上不可預測。他另曾探討 AGI 之後的勞動經濟學（Hutter 2026），並在報告討論刻意減速時獲得引用；他也研究過棲居於運算虛擬世界中的數位智慧（Hutter 2012）。

相關連結#

待解決的問題#

  • AIXI 無法計算，也不具嵌入性；近期的改進（攤銷預測器、嵌入式／多代理 AIXI）能在多大程度上讓理論更切合真實 ASI 的實際需求？

資料來源#

  • From AGI to ASI — 共同作者；報告引用 Hutter (2005)、Hutter et al. (2024)、Hutter (2012, 2026)
§ end
Cited by 5
  • Google DeepMind×2

    It is also the source of the wiki's theory-of-superintelligence cluster. The June 2026 report From…

  • Shane Legg×2

    Shane Legg is a co-founder of DeepMind and a long-standing theorist of machine intelligence. With…

  • Entities — People, Orgs, Tools & Projects

    Marcus Hutter — Creator of AIXI and the Universal AI framework; DeepMind senior researcher and ANU…

  • Open Questions Backlog

    Marcus Hutter (106d) — AIXI is incomputable and non-embedded; how far do recent fixes (amortized…

  • Universal AI (AIXI)

    Shane Legg / Marcus Hutter — the originators of the framework and the intelligence measure

Related articles
  • Shane Legg

    Co-founder and Chief AGI Scientist of Google DeepMind; co-author with Hutter of the Legg–Hutter universal intelligence…

  • AGI-to-ASI Pathways

    DeepMind's four non-exclusive, parallel technological routes from human-level AGI to superintelligence — scaling, algor…

  • Artificial Superintelligence (ASI)

    DeepMind's informal characterization of ASI as a system that exceeds large, well-coordinated human-expert collectives a…

  • Universal AI (AIXI)

    Hutter & Legg's formal upper bound on machine intelligence: AIXI, the incomputable agent optimal on average over all co…

  • The Abstraction Barrier

    Lerchner's hypothesis that AI trained on human concepts may be unable to discover genuinely novel conceptual primitives…