The problem of induction: a turkey cannot possibly predict what happens on the afternoon before Thanksgiving.
Statements presented as science may not be as rigorous as they seem. Finding confirmation is easy.
We are suckers for narratives. Hindsight attributes causality where none was apparent, e.g. the series of events that led to WWII. However, negative confirmation from history is invaluable.
Some live in the antechamber of hope, waiting for the big break. But VCs do better entrepreneurs, publishers over writers, dealers over artists, etc.
The problem of silent evidence. How many courageous, risk-taking, optimistic, etc. people never become millionaires? Compute the odds based on the starting cohort, not the ones that succeed.
The Uncertainty of the Nerd: Why may a street fighter beat an Olympic wrestler? Because the wrestler has been trained to ignore what the rules disallow, e.g. kicks to the groin.
But if we observe all turkeys in the US, won’t we notice that they’re more likely to die right before Thanksgiving? Zooming out to the bigger picture invalidates Taleb’s point.