Reader Guide

One-line summary

On Gradual AGI develops an integrated framework for examining gradual relations between AI capability and realization, as mediated by participation, institutions, and evidence, within AGI-inclusive humanity, under an adopted commitment to the benefit of human beings situated within it.

Principal Contributions

The book makes the following principal contributions. Originality lies in their formulation, relationships, and integration; established methods and antecedents are credited in the chapters.

1. Gradualness, graduality, and relational realization. The book distinguishes phenomenal gradualness from principled graduality and capability potential from realization for specified groups under specified conditions. It places these distinctions within AGI-inclusive humanity, under an adopted commitment to the benefit of human beings situated within that configuration (Chapters 1–2).

2. A Multi-Plateau Framework. The book provides an architecture for distinguishing, relating, and revising different kinds of claims. Its plateaus are logical vantage points, not chronological stages or levels of AGI development. Chapter 3 formulates the architecture of epistemic repairability, including dependency-based propagation, version preservation, and repair without rescue.

3. A group-indexed realization mapping. The mapping relates realization to four arguments: Potential, Availability, group-specific Exposure, and domain-specific conditions. It distinguishes capability from the conditions under which capability becomes realized, without making benefit an additional argument (Chapter 4).

4. A synchronization model with bounded applicability. Ceiling–Floor–Slope (CFS) specifies the evolution of a capability–realization gap where its applicability conditions hold. Its contribution lies in the formulation, interpretation, and integration of the model, including convergence and divergence regimes, gap-dependent rates, delay, and an explicit applicability stop rule, rather than in claiming new recurrence mathematics (Chapter 5).

5. Distinct roles for Optimization, governance, and Participation. The book relates trajectory evaluation, contestation, and participatory processes to realization while preserving their different analytical roles. Optimization remains upstream of the realization mapping; governance and Participation do not become extra arguments of it (Chapters 6–8). Chapter 7 proposes specialized critic models within accountable contestation, tracing a 2025 proposal; its equilibrium interpretation and comparative effectiveness remain untested.

6. Participation under empirical contact. Chapter 9 brings Chapter 8’s human-side conceptual apparatus, situated within reciprocal human–AI Participation, into empirical contact and distinguishes discriminant evidence, observed sequences, unresolved reliability burdens, and outcomes that remain unadjudicated.

7. Epistemic discipline and differentiated feedback. Chapter 10 separates underlying state, observation, representation, inference, and claim status, and applies and develops Chapter 3’s repair architecture in the evidence setting. Chapter 11 distinguishes temporal return, feedback, and recursion without treating every sequence as recursive self-improvement.

8. Provisional architectural integration. Chapter 12 brings these components together while preserving their boundaries, applies the architecture across Go, mathematics, and radiology, and provides a practical workflow for evaluating claims. It introduces no new apparatus and does not claim a closed or final system.

How the book develops

The book moves through four Parts. These are reader-level movements of inquiry rather than chronological stages, stages of AGI development, or the Multi-Plateau Framework’s plateaus.

Part I, Orientation (Chapters 1–2), establishes the meaning of AGI, phenomenal gradualness, principled graduality, and AGI-inclusive humanity. It separates descriptive claims from the adopted normative commitment to human benefit.

Part II, Architecture (Chapters 3–6), develops the Multi-Plateau Framework, the realization mapping, the synchronization model, and the evaluation of trajectories. Optimization is upstream of the realization mapping, not an argument of it.

Part III, Mediation (Chapters 7–11), examines human, institutional, and temporal processes that shape realization, together with the evidence and epistemic discipline through which claims about it are assessed. Governance and Participation are brought into empirical contact, followed by the treatment of observation, feedback, and recursion.

Part IV, Integration (Chapter 12), brings the components together provisionally, without totalization—that is, without claiming a closed or final system. It also contains the Glossary, Full Notation Guide, and compact Claims and Status Register. “Compact” describes the register’s reader-facing purpose, rather than an abbreviated version of an otherwise identical longer register.

Prerequisites and reading routes

No prior acquaintance with the book’s formalism is assumed. Chapters 4–6 require particular attention to notation and declared conditions; the early Notation Quick Reference supports that reading. Readers can begin with the conceptual orientation, then work through those chapters’ formal specifications alongside their explanations.

Sequential route. Read Chapters 1–12 in order to follow the cumulative argument.

Architecture route. Read Chapters 1–6, then Chapter 12; return to Chapters 7–11 for the processes and evidentiary disciplines through which the architecture is applied and revised.

Governance and Participation route. Begin with Chapters 1–2, consult Chapter 3 for logical type and evidentiary standing, and read the realization mapping in Chapter 4 and the evaluation–authority boundary in Chapter 6 before Chapters 7–9. Chapters 10–11 clarify the evidentiary and temporal burdens of claims made along this route.

Overview route. Begin with Chapter 12, which is designed to be readable as a stand-alone overview. It compresses distinctions developed in earlier chapters, so readers beginning here should expect to look back for their full arguments, formal specifications, and evidence.

Each chapter’s Reader Guide provides a one-line summary, identifies its principal contributions, states prerequisites and reading routes, and explains its evidence and principal limits. Separate abstracts and keywords are retained where present in the chapter sources; every chapter has a Reader Guide.

Design principles: Open, Flexible, Falsifiable

Open: the framework invites participation, makes its architecture inspectable, and supports traceable changes. Flexible: it remains available for update, revision, and repair. Falsifiable: it commits to formulating propositions and claims for empirical testing wherever the domain permits, specifying what would count against them.

These design principles operate throughout the book. Orientation makes assumptions explicit; Architecture makes relationships inspectable; Mediation brings claims into contact with institutions, participation, evidence, and time; Integration exposes the relationships and limits of the resulting model. The principles are not assigned exclusively to particular Parts.

These principles do not imply that every conceptual distinction or normative commitment is an empirical hypothesis, that all propositions are presently testable, or that finite non-refutation establishes truth.

Evidence and limits

The book distinguishes conceptual formulation, formal specification, illustration, and empirical finding. A model specifies relationships under declared conditions; an illustration shows how a distinction can be used; a finding carries only the evidentiary standing earned by its design and sources. Claim status remains local to the proposition, domain, and available evidence.

The book does not claim universal empirical validation of its architecture, that capability guarantees realization or benefit, or that observed feedback establishes autonomous recursive self-improvement. Introducing the distinctions between phenomenal gradualness and principled graduality does not establish that their applications have been empirically validated.

Supporting materials

The early Notation Quick Reference supports first encounters with formal expressions. Part IV provides the Glossary, Full Notation Guide, and compact Claims and Status Register. The Canonical Extraction / Provenance Ledger is provided separately as scholarly back matter and records the source and revision history supporting the assembly. Unified References provide the book’s authoritative reference list.

The framework is offered for use, testing, criticism, and revision. Its value depends on whether its distinctions help readers make better claims about what becomes realized, for whom, under which conditions, and on what evidence.