roll comments on How can we ensure that a Friendly AI team will be sane enough?

[roll]

Of course, but one can lack information and conclude “okay, I don’t have enough information”, or one may not arrive at such conclusion due to the overconfidence (for example).