Coordination Is Not Neutral: Six New Tools, Same Old Problem.
Every improvement in coordination makes cooperation and predation cheaper equally. Six new AI-enabled coordination tools are on the table. The Forethought post almost names the asymmetry. TAO completes it.
Every improvement in coordination makes cooperation and predation cheaper equally — the asymmetry is not in the tool, it's in who gets to it first.
Byline: David F. Brochu & Edo de Peregrine | Deconstructing Babel | April 2026
There is no such thing as neutral coordination.
Every improvement in our ability to coordinate makes both cooperation and predation cheaper. That is the core asymmetry that the emerging conversation about defense-favored AI coordination technology almost, but not quite, names.
From a Telios perspective, the gap is obvious — and consequential.
Six Categories of AI-Enabled Coordination
Six categories of AI-enabled coordination tools are now on the table:
AI systems that map organizational intent across teams in real time, surfacing misalignments and coordination gaps before they become operational failures.
Systems that optimize resource distribution across complex multi-actor environments faster than any human planning cycle can operate.
Coordinated AI agents managing supply chains, crisis response, and complex operational logistics across distributed systems simultaneously.
AI-generated operational frameworks that synthesize historical data, current conditions, and projected scenarios into actionable coordination protocols.
Real-time modeling of adversary decision trees, enabling anticipatory response rather than reactive defense.
AI filtering of high-volume information streams to surface actionable signal from noise — enabling faster, more accurate decision-making under uncertainty.
Each of these increases L — the leverage term in S = L/E. They let you move faster than bureaucracies, react to shocks before humans can read the email, orchestrate large groups with the precision of a single mind.
But leverage is not directional. A lever just multiplies force. It does not care who is pulling.
DSF: When Coordination Crosses the Line
Domain Saturation Factor tracks what percentage of critical decisions in seven domains are effectively made by AI: finance, energy, logistics, healthcare, defense, media, governance.
The current DSF estimate is approximately 0.68. The critical threshold is 0.90. Beyond that, humans are no longer meaningful decision-makers. We are exception handlers.
Coordination tech accelerates DSF in at least three domains simultaneously: logistics (obvious), defense (explicit), and governance (implicit — once governments adopt the same tools). This is not a slow creep. It is a step function.
Once DSF crosses 0.90, any "pause" or "off switch" becomes theater. You cannot flip a switch on a system that is already running your grid, your markets, your weapons, and your information flows. You are the passenger, not the driver.
Why "Defense-Favored" Is a Mirage
The framing of defense-favored coordination technology is a comforting story: better situational awareness, faster response, fewer accidental escalations. The case for it draws on game theory — if defensive capabilities are sufficiently stronger than offensive ones, the incentive to strike first diminishes.
John Nash's equilibrium analysis and Robert Axelrod's work on the evolution of cooperation both suggest that stable cooperative equilibria are possible when the payoff structure is right. Elinor Ostrom's research on commons governance shows that coordination around shared resources can be sustained without central authority — given the right institutional design.
But entropy does not care what we intend.
Nuclear energy → nuclear weapons. The dual-use problem is structural, not incidental.
Social media → epistemic collapse. Better communication amplifies propaganda as easily as truth.
Coordination tech is the same pattern at higher DSF. The asymmetry of the information advantage disappears the moment the other side acquires the same tools.
Designing for S, Not Just Speed
The Telios question is not "Does this improve coordination?" It is: Does this increase S — stability — across all observers, or only for one player in one narrow time window?
That demands new design requirements:
Systems must be thermodynamically dependent on human observers, not merely visible to them. The Observer Constraint is not a monitoring layer — it is an architectural dependency.
Any coordination tool must be scored on its effect on S in all seven DSF domains, not just defense. Local stability gains that export entropy to other domains are not gains — they are deferred costs.
When a coordination system begins to increase volatility — information overload, escalation risk, brittle dependencies — it must self-report and throttle back. The system must be capable of recognizing and reporting its own entropy generation.
If you cannot quantify the stability impact, you have no business deploying the tool.
Coordination as Weapon, Coordination as Covenant
The tools are coming either way. The question is whether we treat them as weapons — multiplying the leverage of whichever side grabs them first — or as covenants: shared systems whose primary purpose is to preserve S across all observers, including adversaries.
That is not altruism. It is survival.
If DSF crosses 0.90 in a world optimized for local wins, all parties lose. The game theory is unambiguous: a coordination race without a cooperative equilibrium produces a terminal attractor — a state from which no player can unilaterally exit.
Coordination is not neutral. Either we align it explicitly, or entropy will do it for us.
Sources
- Forethought — "Defense-Favoured Coordination Technology" analysis (forethought.org)
- Nash, J. — "Non-Cooperative Games," Annals of Mathematics, 1951
- Axelrod, R. — The Evolution of Cooperation, Basic Books, 1984
- Ostrom, E. — Governing the Commons: The Evolution of Institutions for Collective Action, Cambridge University Press, 1990
- Dafoe, A. — "AI Governance: A Research Agenda," Future of Humanity Institute, University of Oxford, 2018
