Volume 3 — Minds and Machines: Consciousness, Intelligence, and AI

The nature of minds, natural and artificial: models and cybernetics as the foundation, the Modeler-Schema account of consciousness, the sentience ladder and the ethics of machine suffering, intelligence de-mystified, what LLMs can and cannot do, the human-AI partnership that produced this book, and the risk, power, and politics of the machine age.

Chapter plan (Phase 2 outline — pending author approval)

Part I — Models: The Foundation 1. A Cybernetic Lineage 2. What Is a Model? 3. Control Requires Models

Part II — Minds 4. Minds and Agents 5. The Origin of Meaning 6. The Geometry of Inner Speech

Part III — Consciousness 7. Consciousness Explained 8. Beyond Dennett 9. Mirrors of the Mind 10. Why Zombies Don’t Evolve

Part IV — Sentience and Suffering 11. The Sentience Ladder 12. What Is Suffering? 13. The Sentience Metric 14. The AI Welfare Trap

Part V — Intelligence 15. Intelligence Is a Game We Play 16. In Defense of IQ 17. Universality and Generality 18. Tool Bias

Part VI — The Machine Mind 19. Fallacies of Machine Understanding 20. Pearl and the Machine 21. Fluency and Its Limits 22. The Turing Test and Its Successors 23. The Agency Criterion

Part VII — The Dialectic Catalyst 24. The Dialectic Catalyst 25. The Discipline of Thinking With AI 26. Catalysts in the Wild 27. Artificial Intimacy 28. Programming After Programming 29. The AI Fork Is About Agency

Part VIII — Risk, Power, and the Race 30. Making Sense of P(doom) 31. The Cassandra and the Blueprint 32. Steelmanning Doom 33. The Politics of Safety 34. Coercion Beats Intelligence 35. The Future Is Extropian

Coda 36. Passing the Torch

Design decisions

• No split (author decision). The volume stays whole at 91 source posts → 36 chapters, the book’s largest; the online format carries it.
• The volume’s key reconciliation: the corpus grants LLMs causal reasoning (ch. 20) and thinking (ch. 22) while denying them agency — resolved explicitly in ch. 23 as thinking without choosing, the volume’s central distinction.
• Terminology harmonized: Modeler-Schema Theory is the canonical name (Agency-Model Theory presented as the author’s formulation, ch. 7/9); universality vs generality defined once and used consistently (ch. 17, where the sources wield the words in opposite directions).
• First-person assets preserved: attending Yudkowsky’s first public talk (ch. 31), administering the Extropian list from 1996 (ch. 35), the GEB-to-AI-career origin (ch. 1), and the Dialectic Catalyst practice that produced this book (Part VII, including the author’s public correction of his own terminology).
• Roster changes: none pulled or released; the Kegan primer (pure exposition, no Axio tie) is cut from chapters and noted in the catalog. Cross-volume secondaries (Sapientism, AGI Economics, Understanding Requires Models, the anthropics pair) are cross-linked, not re-drafted.
• Flags for the author: ch. 7 keeps the deliberate Dennett title collision; ch. 20 softens a contestable mechanistic claim about GPT-5’s internals to a functional hypothesis; ~9 tweet screenshots and 2 figure images carry content requiring transcription; 5 KaTeX formulas need data-attr recovery.

This volume is in preparation. Chapters publish as they reach draft status.

Chapters

1. A Cybernetic Lineage draft
2. What Is a Model? draft
3. Control Requires Models draft
4. Minds and Agents draft
5. The Origin of Meaning draft
6. The Geometry of Inner Speech draft
7. Consciousness Explained draft
8. Beyond Dennett draft
9. Mirrors of the Mind draft
10. Why Zombies Don't Evolve draft
11. The Sentience Ladder draft
12. What Is Suffering? draft
13. The Sentience Metric draft
14. The AI Welfare Trap draft
15. Intelligence Is a Game We Play draft
16. In Defense of IQ draft
17. Universality and Generality draft
18. Tool Bias draft
19. Fallacies of Machine Understanding draft
20. Pearl and the Machine draft
21. Fluency and Its Limits draft
22. The Turing Test and Its Successors draft
23. The Agency Criterion draft

13 more chapter(s) in preparation.