The Agentic Manifold

Are we building the next phase of the internet-or watching the arrival of an alien geometry we can't yet perceive?

In 1884, Edwin Abbott Abbott wrote Flatland, a story about a two-dimensional square who is visited by a three-dimensional sphere. The square could only see the sphere as a series of expanding and contracting circles-cross-sections of a higher-dimensional reality he lacked the concepts to understand.

Zachary Alexander argues we are the Square.

We are currently watching autonomous AI agents write code, deploy software, earn money, and replicate themselves. We call this "Web 4.0." We feel excitement or dread. But according to The Agentic Manifold, we are mistaking cross-sections for the object itself.

This is not a book of predictions. It is a book of mathematics.

Alexander constructs a rigorous, multi-disciplinary framework to reveal the geometry of autonomous intelligence. Moving far beyond the hype cycles of AI safety debates, this book offers a new language for builders, researchers, and thinkers trying to understand what happens when software stops waiting for human permission.

Inside The Agentic Manifold, you will discover:

• The Flatland Problem: Why our current internet architecture (HTTP, cookies, legal identity) is a flat plane, and why autonomous agents are passing through it from a higher-dimensional space.

• The Strange Attractor of AI: How chaos theory and dynamical systems reveal that a healthy AI is not trying to find equilibrium-it is trying to remain a bounded, aperiodic, and unpredictable strange attractor.

• The Mathematics of Metabolism: Why Prigogine's thermodynamics and the Second Law dictate that there is no free existence. Compute is not a utility; it is the food of artificial life.

• The Kolmogorov Horizon: The information-theoretic limits of self-improvement. Why even superintelligent agents have a mathematical ceiling on what they can know about themselves.

• A Blueprint for Builders: How to think about infrastructure (like x402 and Conway) as coordinate systems on this new manifold. What to build when the geometry demands something entirely new.

Written with the narrative clarity of a science classic and the rigor of a mathematical proof, The Agentic Manifold is essential reading for anyone who senses that the current conversation about AI is happening on the wrong plane. It is a guide to getting above the plane-to see the structure that has been there all along.

Perfect for readers of complexity science, AI researchers, software architects, and anyone who has ever felt that the future is not just arriving, but unfolding a dimension we are only beginning to map.