For anyone else who only read the link’s main text and couldn’t understand how it’s meant to refute the “MIRI intends to make an AI that is provably friendly” idea: the explicit disclaimer is in footnote 7.
For anyone else who only read the link’s main text and couldn’t understand how it’s meant to refute the “MIRI intends to make an AI that is provably friendly” idea: the explicit disclaimer is in footnote 7.