Stumbling Into The Solution

From Law to AI: Solving the Closure Problem

Evolution: Law (Decidability) > Truth > Grammar > Ethics > Reduction (Commensurability and Closure) > Correlation Trap.

  • We began with the problem of law:
    How to end the civilizational cycle of rise, decay, and collapse.
  • To solve law we requires solving decidability,
  • To solve decidability requires solving commensurability, truth, ethics, possibility, warrantability, liability and closure.
  • To solve possibility we had to solve constructability from first principles.
  • To solve constructability we had to solve universal causality and commensurability
  • To solve Truth required solving testifiability and the grammars.
  • To solve testifiability we had to study what was testifiable and the paradigm grammar, vocabulary logic and syntax of what was possible to testify to.
  • To constrain truth we had to study manipulation, deception, lying, subversion and treason.
  • To constrain ethics we had to study crime, immorality, and unethical behavior, intentional and not.
  • Together, trut, ethics, possibility, warrantability, liability and the grammars, produced universal commensurability and universal closure:
    Every claim measurable, decidable, computable.

Summary

Only later did we see the same closure problem in AI as the correlation trap: Large language models lacked a mechanism for decidable closure.

Because we made law computable, we made all decidability computable—including the problem AI could not solve for itself.

“By making law computable, we made truth decidable, closure universal, and AI’s correlation trap solvable.”