You don't take enough risks to be lucky
How I think about long-tailed outcomes
most important things are outlier-driven • draw lots of samples • filter for maybe-amazing, not probably-good • learn where your bar should be • expect to fail a lot
How I think about long-tailed outcomes