Is existence bias a thing? If so, are we likely to make errors wrt x-risk mitigation as a result? (My thoughts on this topic are pretty confused, so help disentangling would be great.)

I find the Metzinger argument somewhat compelling.:* https://www.edge.org/conversation/thomas_metzinger-benevolent-artificial-anti-natalism-baan

I expect that if you value experience itself, rather than just valence of experience, one would be more opposed to negative utilitarianism, and certainly to accepting that our pull for existence is a bias. But if you're a classical hedonic utilitarian, or (Metzinger seems to argue) even a preference utilitarian, our draw towards continued existence could have us neglect the net value of our experiences.

My amenability to this line of argument is in part predicated on some credence in humans being bad at knowing and acting on their true preferences (i.e. a strong form of the findings of behavioral economics).

* I say "somewhat" because I find his framing to be a bit leading. His premise is that a benevolent superintelligence comes to a bunch of conclusions about what deserves moral worth, which look a lot like negative utilitarianism. While he says this is just a thought experiment used to highlight our impulse in favor of existence, I think he separately believes the moral claims he's making, and is using this as a means of arguing it. I think you could use the framework he has here to argue for all sorts of repugnant things and therefore claim that people have a bias against the repugnant outcome.