Refactoring cryonics as structural brain preservation — LessWrong