The Final Laws of Geopolitical Game Theory
Chokepoints, AI, Empire, and the Hidden Logic of Power — Lessons from Professor Jiang
Book Two of The Laws of Game Theory
Volume One gave you twenty laws. This one completes the system — and then shows you how the pieces fit.
The second semester is the harder half. It opens by refusing the explanation every newspaper reached for, and reads a published defence strategy line by line instead, translating each euphemism back into the operation it authorises. From there: how a population becomes the weapon that destroys its own government, why grand strategy is a civilisation's story about itself made operational, and what a refusal to define an objective tells you about the objective.
The seven laws
- The Law of the Chokepoint — an empire does not rule territory; it rules the narrow places through which everyone else must pass, and it keeps its rivals paired off so all of them remain customers.
- The Law of Strangulation — when you cannot destroy the army and cannot destroy the population, you make the population the weapon that destroys its own government.
- The Law of Grand Strategy — strategy is culture made operational, which makes it legible in advance from a civilisation's own canon, and impossible to discard when it fails.
- The Machine God — when a powerful actor refuses to define the objective it claims to pursue, the refusal is the strategy: an undefined goal can never be audited or declared complete.
- The Law of Derivative Power — a power constituted by another power cannot defeat it. Dependence sets the ceiling.
- The Law of Coordination — a minority that can coordinate across domains beats domains that cannot coordinate with each other, and its advantage grows as institutions decay.
- The Law of the Junior Partner — closeness is not leverage, because the senior partner's real dependency lies elsewhere.
What makes this volume different
Every chapter closes with a Game-Theory Instrument: the players, the payoffs, the constraints, the dominant move, the equilibrium, and what would change the game. You can lift that block out and apply it to something in your own week.
And the book ends with an appendix that very few books of this kind contain — a statement of what would prove the whole framework wrong, including where its author thinks it probably overreaches. A model that cannot fail cannot inform.
The closing chapters
The final sessions are a question-and-answer marathon in which the professor is examined by his own students, concedes that he oversimplifies, names three specific outcomes that would refute him — and then announces that he is leaving teaching after four years.