Calculating Bayes rule for everything can be quite weird for a lot of people. I remember a case where someone found it weird that another person asked on LW how to do a Bayesian calculation for the likelihood that a specific girl likes him.

Calculating probabilities for many everyday issues is hugely weird for many people. You might even have to take care to make it sound believable even if you do describe a real world character.

I remember an anecdote of a person doing an utility calculation that suggest having sex without a condom and being exposed to the chance of getting AIDS is quite okay.

Another of those things that CFAR preaches that can be seen as pretty weird is purposeful comfort zone extension. It's the kind of topic where you also have to worry about believability if you just tell real world stories.