And to be clear, this largely makes sense: humans are not good at preventing long tail failures that require constant vigilance and are not responsive to unlike

And to be clear, this largely makes sense: humans are not good at preventing long tail failures that require constant vigilance and are not responsive to unlikely punishments, no matter how severe. The thing we have to do is improve the systems in which people operate.