How to Read This Book | AI Wars of 2026
A guide to the argument and its evidence. A page from the AI Wars of 2026 research edition.
How to Read This Book
This book examines AI as a set of overlapping strategic contests, not simply as a leaderboard of models. Models matter, but so do agents, applications, platforms, clouds, devices, data, identity, authority, execution, and the workflows where people decide whether to trust an AI system.
Start with Part 0 to understand the contests, prizes, and tests of winning. Part I develops the shift from answers to delegated action. Part II applies that frame to core and provisional contenders with different starting positions. Later research sections examine the systems and battlegrounds that can change the balance of power.
Read the claims with their boundaries
Each chapter begins with a Big Idea: a specific proposition that matters strategically and can be tested. The chapter then identifies what the evidence supports, what it does not, and what could change the conclusion.
Source numbers open the records behind factual claims. Those records identify who reported the evidence, when, and the limits of what it can establish.
This is a living book
The research record changes as better evidence appears. Pages marked New were recently added. Pages marked Updated have changed since you last completed them. Pages marked Read are completed at their current version. A page returns to Updated when a substantive revision changes what you previously completed.
The book follows more than model releases. It asks how intelligence becomes part of work, who earns the right to act on a user's behalf, and where the resulting value may accumulate. A product announcement establishes what a company has introduced or claimed; it cannot independently establish market share, productivity, reliability, or a lasting competitive advantage.
Treat the conclusion as provisional
The purpose is not to celebrate a technology cycle or crown a company early. It is to make the underlying strategic question inspectable:
Which positions turn intelligence into trusted, consequential action—and what evidence would show they create durable value?