Our instinct in the knowledge work setting is to obsess about factors like worst-case scenarios—how can we prevent bad things from ever happening?!—or to prefer the convenience of simple (but costly) protocols to more finicky (but optimized) alternatives. The information theory revolution tells us that these instincts shouldn’t be trusted. Take the time to build the protocol that has the best average cost, even if it’s not the most natural option in the moment, as the long-term performance gains can be substantial.
