Its also just bad understanding of probability. Making bets to end the world isn't a one time deal, its an obvious repeated game. You don't just need to win one 50-50 bet, humanity needs to win all of them which approaches 0 probability.
The focus on "rational" value maximization on individual bets is doubly blind:
- First, you and I might have different utility functions which are based on valuing outcomes differently; if utility functions are different, there may not be a single "most rational" choice
- Second, a single bet is a single bet, but investing in a company or following a specific policy is more like staking a gambler to make a sequence of bets.