Act 3 — MMM

The third act names MMM — Measure and Map, Model and Manipulate, Manifest and Make — as the most meta process we can observe: the same three-layer shape recurs every time someone tries to describe what agency or intelligence even is, from single-cell life to civilisation to contemporary AI research.

Overview

MMM is the book’s answer to the question: given that beings want (Act 1) and draw distinctions (Act 2), how do they grow? The answer is a recursive three-step loop that follows an ontological → relational → generative arc: Measure/Map (discriminate what pieces exist; chart the territory), Model/Manipulate (build a causal model of how pieces connect; probe its behaviour to discover how things work), Manifest/Make (synthesise something new from discovered components and rules). The loop runs at every scale — in molecular biology, in learning, in evolution, in science, in contemplative practice, in daily life.

The act’s central claim is not that MMM is a recipe but that it is the most meta process we can observe. The evidence is structural convergence. Every serious attempt to describe agency or intelligence across disciplines produces a three-layer shape. A 2026 paper by Emmanuel Dupoux, Yann LeCun, and Jitendra Malik (“Why AI systems don’t learn and what to do about it: Lessons on autonomous learning from cognitive science”, arXiv:2603.15381) proposes a three-system architecture for autonomous intelligence: System A (observation learning — passive, watching the world), System B (active-behavior learning — interacting, probing, discovering causal structure), and System M (meta-control — switching between A and B based on context). This is a structural echo of Measure, Model, Manifest. The book does not claim the match is exact; it is structural, not detailed. The point is that when the three-layer shape recurs across independent attempts to describe agency — from cognitive science, from AI research, from the wisdom traditions, from the biology of life itself — it is probably tracking something real about what being an agent requires.

The act’s eleven chapters develop the loop in four phases. Chapter 12 opens the act by mapping the information realm — the shape of knowledge that the loop operates on. Chapters 13–15 introduce the three moves (Measure, Model, Manifest) in isolation, with enough precision that they can be distinguished when one is stalled. Chapter 16 reassembles them as the lived experience of growth, showing that the loop is not a metaphor borrowed from one domain but the same shape running on different substrates, and names this as the most meta process we can observe. Chapters 17–18 turn the frame outward: chapter 17 shows how the info realm folds back into action (applied knowledge), and chapter 18 treats how we change the world through the loop’s outputs. Chapter 19 extends the argument to life itself — the loop runs in things without brains, civilisation is recursive MMM loops stacked on top of biological ones, and AI research keeps rediscovering the same shape. Chapters 20–21 apply the frame: chapter 20 to wisdom traditions (read as MMM-shaped practices), chapter 21 to practical concerns (self-esteem, trust, relationships, money, meaning). Chapter 22 closes the act — and the book’s argument — by asking what objectivity looks like for finite beings running the loop: disciplined approximation, not omniscience.

The intellectual inheritance here is broad: Boyd’s OODA loop; Kolb’s experiential learning cycle; Bateson’s learning hierarchies; Argyris and Schon’s single-loop vs. double-loop distinction; the free-energy principle (Friston); and the Dupoux/LeCun/Malik (2026) A/B/M architecture as contemporary independent convergence from AI research. The act insists the loop is not just a useful heuristic but the actual shape of growth across substrates.

Chapters in this act

Key claims

• MMM is the most meta process we can observe: every time someone — in biology, psychology, or AI — tries to write down what it takes to be an agent, a three-layer shape falls out of the attempt. — Source: /src/content/chapters/16-mmm-as-lived-experience.md, /src/content/chapters/19-mmm-across-life-itself.md
• Independent convergence: Dupoux, LeCun, and Malik’s 2026 three-system architecture for autonomous intelligence — observation learning (System A), active-behavior learning (System B), meta-control (System M) — is a structural echo of Measure / Model / Manifest. The match is structural, not detailed; the significance is that the same shape recurs across disciplines. — Source: /src/content/chapters/16-mmm-as-lived-experience.md
• Growth is the felt experience of the MMM loop turning: better measurement, richer models, more effective manifestation. — Source: /src/content/chapters/16-mmm-as-lived-experience.md
• The loop can stall at any of its three points; most felt suffering is a stalled loop — movement returns when the missing move is identified. — Source: /src/content/chapters/16-mmm-as-lived-experience.md
• Every measurement instrument carries a theory of what matters; there is no neutral measurement. Measuring is not modelling — the map is a census of what exists before any causal account. — Source: /src/content/chapters/13-measure.md
• The manipulate phase is what distinguishes a real model from a static hypothesis: you know you understand something when you can move its parts and predict what changes. — Source: /src/content/chapters/14-model.md
• A model is a story so disciplined it is willing to be wrong out loud; better living often means better modelling, not more effort. — Source: /src/content/chapters/14-model.md
• Manifest is genuine creation — bringing new things into the world from discovered components and relationships — not merely action or wager. Without making, the loop stalls and knowledge remains inert. — Source: /src/content/chapters/15-manifest.md
• Civilisations are recursive MMM loops running on top of biological MMM loops. — Source: /src/content/chapters/19-mmm-across-life-itself.md
• Wisdom traditions converge on a recognisable triad: discipline of attention (Measure), ontology (Model), practice or ethic (Manifest). — Source: /src/content/chapters/20-what-the-mystics-were-tracking.md
• Objectivity for finite beings is not omniscience but discipline: honour reality, mark uncertainty, distinguish levels, track consequences, stay revisable. — Source: /src/content/chapters/22-the-most-objective-we-can-be.md

Connections

• [[The MMM Loop]] — the central concept; full entry with detailed breakdown
• [[Growth and Learning]] — thematic treatment across Act 3
• [[Wisdom Traditions]] — chapter 20’s argument synthesised
• [[Objectivity]] — chapters 4 and 22 form a bracket; full treatment in the theme page
• [[Act 1 — Volo Ergo Sum]] — the loop requires want as its engine
• [[Act 2 — init.d]] — the loop requires distinctions to have something to measure
• [[Thingification]] — institutions built by MMM loops are the most consequential thingifications

Open questions

• The conditions under which the loop starts and stops are “partly open” — the book does not give a full account of what kicks a stalled loop back into motion.
• The right ratio of Measure to Model to Manifest is explicitly left domain-dependent; no general theory offered.
• Why the loop runs at all — what makes matter start striving — is flagged as an open question inherited from Act 1.

Last updated: 2026-04-16 | Sources: 8 — framing sharpened 2026-04-16 (meta + Dupoux/LeCun/Malik)