I agree with those who believe we should do what we legally can to prevent the emergence of unaligned agentic superintelligence. I also agree with those who believe humanity should disperse in multiple and random directions as soon as possible, with Mars as a secondary base. Yet questions about strategy remain for those of us who do not have the opportunity to disperse. It seems that, as humans continue to roboticize and digitize themselves, the line between species will become increasingly blurred. Furthermore, gradual merges, in which one leaves a backup copy of oneself but chooses to slowly take on the intelligence and memories of other individuals (thus becoming gradually more enlightened) would be mutually advantageous. These new beings will most likely run simulations, as has been described elsewhere. However, there will be no reason to run second-hand simulations on separate hardware, as humans currently do. Rather, such merged beings could simply unidirectionally and temporarily blind part of themselves for this activity, and thus directly gain additional enlightenment. As others have pointed out, since a single future being will be able to run many simulations, we are most likely in such a simulation now. If the simulation is on a self-hosting substrate, which would be to the advantage of the host, we would furthermore be unaware of our true identity and greater self.