TriadigmResearch Orientation
Front cover of Governance That Learns: Accountable Power in the Age of AI, by Andre Milchman.

Web edition · Read in full

Governance That Learns

Accountable Power in the Age of AI

Andre Milchman

A public manifesto for governing artificial intelligence: develop the capacity to govern as deliberately as we develop the technology. Andre Milchman asks what governance should make possible, what it must protect, and how we will know when it needs to change.

Read the book Buy on Amazon Table of contents →

The complete book — introduction, three parts, twelve chapters, conclusion, four appendices and endnotes — is published here and readable in full. Print and ebook editions are available on Amazon.

Contents

The complete manuscript.

Numbered references in the text link to the endnotes; each endnote links back to the passage that cites it.

  1. IntroductionIntroduction
  2. Part IPurpose and Power
  3. 1What Is Governance For?
  4. 2What Are We Actually Worried About?
  5. 3What Must Governance Be Able to Do?
  6. 4Who Gets a Say—and Who Decides?
  7. Part IIKnowing Enough to Act
  8. 5How Do We Know What Is Happening?
  9. 6How Can AI Help Us Govern?
  10. 7When Should We Act—and When Should We Not?
  11. 8How Do We Act Together?
  12. Part IIILearning Without Losing What Matters
  13. 9What Must We Be Free to Choose?
  14. 10What Are We Assuming?
  15. 11How Does Governance Learn?
  16. 12Who Governs the Governors?
  17. ConclusionMore Capable, Still Answerable
  18. Appendix AThe AI Governance Manifesto
  19. Appendix BThe Governance Amplifier Questions
  20. Appendix CA Governance Assumptions Register
  21. Appendix DThe Governance Frontier: Governance Beyond Human Scale
  22. EndnotesEndnotes
  23. Further ReadingFurther Reading
  24. About the AuthorAbout the Author

About this book

Governance That Learns argues that we must improve our capacity to govern as deliberately as we improve the technologies being governed. It is a public manifesto, not a technical manual: a set of principles, questions, and directions for action addressed to anyone with a stake in how artificial intelligence is developed and used.

The book connects three subjects that belong together — the governance of AI, the use of AI for governance, and the governance of AI-enabled governance — and asks what institutions can understand, what they are entitled to decide, and whether their decisions can change what happens.

Related work on sociotechnical order and responsibility appears in Ecosystems Thinking and Unit Oriented Architecture.