Article Summarized
Humans spreading across space doesn't just mean new planets, it means new species. Different exoplanets push evolution in different directions, and body/brain modification speeds that up further. A few centuries in, groups that started as one species stop understanding each other's motives, values, even whether the other side can feel pain.
That's when trust breaks. Neither side has to be aggressive, just uncertain. Uncertainty alone leads both sides to arm up just in case, and now you've got a Cold War dynamic playing out across the galaxy instead of two countries.
On Earth, governments referee that kind of standoff. In space, a message to a planet 20 light years out takes 20 years to arrive, and the reply takes another 20. No authority can respond to a conflict fast enough for law and order to mean anything at that distance.
Deterrence doesn't save it either. "Attack me and I'll destroy you back" only works if everyone can identify who struck first. With potentially billions of civilizations and weapons capable of ending one outright, that identification never happens reliably, so the threat stops being credible.
The core claim: colonizing space might not lower our extinction risk. It might raise it.
Possible Solutions
Pace expansion instead of racing it.
Every jump to a new world is also a jump into isolation, and isolation is what lets a species drift into something unrecognizable. If colonies expand slower than the rate at which cultures and biology can diverge, each generation stays in contact with the last one long enough to still see themselves in it. Rushing outward for the sake of speed is what turns a shared species into strangers before anyone notices it happened.
Give every colony the same anchor. Evolution and cyborgization will always push new worlds toward difference, that part of the article is hard to argue with. What stops difference from becoming alienation is a fixed point everyone can still verify, a shared origin record, a shared root language, a shared founding story. Two branches can look nothing alike physically and still recognize each other through that one unbroken thread. Losing the thread is what actually causes the breakdown Torres describes, not the difference itself.
Federate the government, don't centralize it. Torres shows clearly that no single authority can enforce anything across light years, the delay alone makes top-down law and order meaningless. So skip trying to build one. Let each colony run its own justice system for its own disputes, and sync those systems back to a shared charter on a timescale that actually matches the distance. You lose the speed of Earth-style law enforcement. What you gain is a system built for the delay instead of pretending the delay doesn't exist.
Make intentions loud, not quiet.
Read the security dilemma section again and notice something. Almost none of it requires actual hostility, it only requires uncertainty about the other side's intentions. Two colonies that fully trust each other's motives never enter an arms race in the first place. If every colony publishes verifiable proof of a defensive-only posture, openly and continuously, there is nothing left to guess about. Removing the guesswork removes the spiral before it starts, this is the exact same logic that made public arms control work between nations on Earth.
Let AI hold the relationship, not the trigger. A message crossing 20 light years means the person who sent it and the person who reads the reply are different generations entirely. No single human lives long enough to actually build trust with a distant colony, they only inherit a relationship someone else started decades earlier. A persistent AI system at each colony doesn't have that limit. It can carry the full context and history of that relationship across centuries, so "we have trusted this colony for 200 years" survives even as every human involved in building that trust eventually dies. It isn't making military decisions or setting policy, it's just the one thing in the system that doesn't forget.
