The Case for Chip-Backed Dollars — LessWrong