We obviously won’t keep the same social contract in a world of ASI; we will get a much worse one. The social contract we have today was premised on the state/ruling class’s need for lots of human beings to make stuff and to shoot guns. This allowed the human beings who supplied these needs to extract concessions over time, like rights to free speech and universal suffrage. But if robots can make all the things and shoot all the guns, then there is simply no basis upon which the vast majority of human beings can negotiate concessions from the ruling class; we will cease to have any kind of leverage at all. At best, this terminates in a kind of permanent techno-feudalism, an infinite future of subjugation under the boot of successive generations of Peter Thiels. At worst, it terminates in a (possibly very slow) genocide of the vast majority of people on earth, because the people who own the robots and the machine intelligences will have no reason to distribute any of the machines’ output to the vast majority of people who lack such ownership stakes. Arguably, our main hope for “survival” in such a future is that the ruling class might have, in this scenario, one final need: a need for subjects to rule over. The psychological need for adoration, for power, for social inferiors. The last thing that we’ll have with which to buy our bread will be supplication to the egos of people like Elon Musk. A future of laughing at a cringe trillionaire’s bad jokes to earn your UBI—more or less directly.