I always find this Yudkowsky-esque conceit somewhat lacking. It's not that one shouldn't examine the motivations of your various... motivations.

But it just seems so close to depersonalization, as witnessed by your interesting pluralization mismatch ('"you" only has'). That supposedly hostile methodology is actually very sensible once you understand it.

I suppose that's what comes from assuming that the hostile hardware is merely a statistical assertion of the Bayesian variety.

And apparently it is functional for some people, so... I suppose that's merit enough.