Colder Wars, By Gwern
TL;DR
Space warfare favors the attacker even more than nuclear war: Gwern maps Cold War first-strike logic onto interstellar conflict and argues all four conditions get worse in space, because attacks can come from any direction, hit with absurd force, resist interception, and arrive with almost no useful warning.
Most science fiction cheats by smuggling old naval warfare into space: He calls out franchises like Star Wars, Star Trek, Battlestar Galactica, and David Weber’s Honor Harrington books for burying implausible assumptions under vivid detail, technical jargon, and emotionally convincing scenes.
Bean’s analysis in Ender’s Shadow gets the problem partly right but still understates it: Card correctly notes that a planet like Earth is a single vulnerable strong point in 3D space, but Gwern says the real attacker would likely skip fleets entirely and use asteroids, comets, or relativistic kill vehicles.
Detection is not the same as defense: Telescopes might spot spacecraft from astonishing distances, even out near Pluto, but seeing an incoming object does not mean you can stop it, especially if it is massive, fast, armored, or actively defending itself with lasers.
MAD breaks down if you cannot identify the attacker: Nuclear deterrence works partly because isotope analysis and launch tracking reveal who struck first, but a body launched from your own Kuiper belt, or stolen from system B and aimed at system C by system A, destroys that accountability.
The Breakdown
Gwern’s core claim is brutal: realistic interstellar war would make the Cold War look tame, because stealthy asteroid or moon strikes are more devastating, less defendable, and far harder to attribute than nuclear attacks. By the end, mutually assured destruction barely looks stabilizing if a solar system can be annihilated by an object launched from its own Kuiper belt and no one can prove who did it.
Was This Useful?
Share
Keep Reading
Make Alcreon Yours
Tune your feedFive quick questions, and the feed ranks what matters to you first.Or just get notified
The weekly Echo. Signal worth keeping in your inbox.
Every new piece, announced on X.
Read Next
See all
Playbook
Cheap Models, Hard Tasks
Most agent workflows route every step to the frontier model by default. The bill scales with how chatty the agent gets, even when most steps don't need that brain.

Playbook
Tasteful Skills
“Tasteful Skills” argues that the best agent skills are not documentation or best-practice lists.

Playbook
The Art of Tasteful Prompting
Learn how tasteful prompting helps you move beyond generic AI output by shaping context, style, and judgment from the start.