In equation form, the AI is maximising

$P\left(\neg X\right)\ast C+P\left(X\right)\ast u(X,A)$

for some constant $C$, some unlikely event $X$ that the AI cannot affect, some set of relevant descriptors $A$, and some utility $u$. Since C is constant, this is exactly the same as maximising $u(X,A)$ - the probability $P\left(X\right)$ is irrelevant.

The whole setup described is simply a way to ensure that if $W$ is the likely set of worlds consistent with observations after $\neg X/X$, then

$P\left(W\right)\approx P\left(\neg X\right)\approx 1$ (we "know" that $X$ doesn't happen and that we end up in $W$),

while

$P\left(W|X\right)<<1$ (in the worlds it cares about, the AI behaves as if $W$ was incredibly unlikely to come about).