However, what I was critiquing was primarily Sotala and Mijic’s arguments in support of technocracy.

The CEV and a FAI might indeed be excellent additions to governance and might naturally rise to the top, but Sotala side-steps that point with the statement “But as you remarked yourself, both in the essay and your comment, the “technocratic” counter-claim is that an AI isn’t a human absolutist ruler, and the comparison is thereby invalid.” That’s special pleading, as is the argument that the CEV will accurately deduce the “real” volition of a populous.

Again, I wonder what the consequences of the CEV determining the extrapolation of a general volition to be impossible, thereby ruling government itself immoral, or some other conundrum.

I think some of the disagreement here may be about how hard it is to specify an AI that does stuff that humans want at least as effectively as humans do already. We consider more-helpful-than-human AI to be possible in the long run, and James and many others don’t seem to. Our argument is never that we should install an AI dictator; merely that it seems personally likely to us that AIs will rise to the positions of higher responsibility because humans choose to vote them into those positions. If that argument makes you uncomfortable, you don’t have to take it too seriously — if people did not want to live in a society where AIs have equal status to humans, then I’m sure that it would still be possible to avoid it. The question that faces us all is this: it seems plausible that superintelligence could be created in the next few decades, and someone will give it some set of motivations, whether it is an enhanced human or AI. The question is “What should those motivations be?” James doesn’t really answer that question, he just complains about connotations of our speaking that he would realize are not about the core issue if he read the source material.