> When you're in a situation where the severity of the outcome doubles every 4 days and you do not have clear information about where the curve will break, underreacting will result in extremely bad outcomes much much much more often than overreacting

> For a simplified example just compare the cost of taking actions that make you break the curve 4 days too early with the situation where you break the curve 4 days too late

But does it seem that implausible that cooler heads might find a solution that doesn't cost at least $20 trillion?