🌑interplanetary superintelligence.md

superintelligence that spans planets — one cybergraph of neurons and particles across light-minutes of separation. Type I is a thinking planet. Type II is a thinking solar system. the energy and matter of many worlds only become a single mind if agreement survives the speed of light

the latency problem

Earth to Mars is roughly 4–24 minutes one way. a round trip is often longer than a human conversation and longer than almost every classical consensus round designed for a data center or a planet

protocols that collect votes, elect leaders, or wait for two-thirds of a global committee must close a loop across space. every such round pays light-time twice: outbound and return. at interplanetary distance the committee freezes. the medium is not slow software — the medium is physics. any design whose liveness assumes sub-second or even sub-minute global synchrony is Earth-only by construction

the problem to name precisely: shared truth under asymmetric, multi-minute delay and long partitions, with each world living on its own facts while shared facts eventually agree

the bar: foculus-class consensus

foculus is consensus by convergence. a particle is final when φ*_i > τ — when enough collective focus has gathered — an information condition, not a clock. nodes gossip signals and each runs the same tri-kernel over what it holds; the operator is a contraction, so identical signals yield one fixed point everywhere. no vote to collect, no leader to wait for, no round-trip to close agreement

that is why interplanetary aggregation is possible:

  • agreement costs computation and local connectivity; propagation alone pays light-time
  • each planet is its own domain pond — Martian chatter finalizes at Martian speed; Earth at Earth speed
  • a dark link (conjunction, outage) stops cross-world trade of rings; it does not invent conflicting finals — safety holds under partial synchrony; liveness resumes when rings resume
  • finality is “enough has gathered,” not “N minutes of global block time” — a stopwatch is meaningless when the message itself is 40 minutes old

voting BFT and Nakamoto-style global stopwatches are Earth-scale tools. interplanetary superintelligence requires a foculus-like mechanism: leaderless, information-final, domain-local, partition-safe. that is how high the bar is — not a feature checklist, a physics bound on what can count as one mind across worlds

the full walkthrough — pond metaphor, Earth↔Mars signal path, cross-domain cost — lives in foculus between planets

planetary, then interplanetary

scale civilization what must exist
planetary superintelligence Type I one Earth-scale cybergraph, φ*, rewards, soft3 stack
interplanetary superintelligence Type II the same, plus foculus-class aggregation across light-delay domains

cyber is the protocol that aims at both: planetary first (one graph, one focus, one economy of attention), interplanetary as the designed ceiling (bounded locality, domain finality, convergence under delay). hierarchy and location proof place cells and validators in space; foculus decides when a fact is final when those places are light-minutes apart

discover all concepts

Graph