Stuart_Armstrong comments on When Goodharting is optimal: linear vs diminishing returns, unlikely vs likely, and other factors

[Stuart_Armstrong]

I like that way of seeing it.