The effect you are mentioning would be magnified for fast running simulations, and slowed down for sluggish slumber ones. So it seems that the relevant unit here is either time-steps, subjective time, markov blankets or some other form of causally constrained model that can be accelerating in beings running at faster speeds.
That said, the effect seems possibly real, and many believe that identity is valuable and determine it’s continuity as a function of the slope of change beings undergo.
The hypothesis you put forth then is that the total value of utilitronium per time-step may be higher depending on how did the universe get to be filled in as densely packet utilitronium, so we should pay attention to the shape and rate of the shockwave, not only to its final state.
The question I’d like to see addressed, assuming you are right, is whether the relevant factors which could compromise or multiply the final absolute value happen before or after the creation of a superintelligence. If after, then the superintelligence itself should decide how to progress.
The discontinuity would increase as subjective time per instant decreased and as subjective change per instant increased. If you have a lot of changes but also a lot of subjective time, it’s fine. So running at faster speeds is fine, as long the number of subjective steps are the same.
Yes, the idea, simply put is that eventually the universe will be filled with utilitronium, and I’m asking whether anything that goes on before that can impact the value of the utilitronium, i.e. the maximal amount of value.
Given that most plausible cross-temporal dependency we know of are on cases of continuity, then the transition from humans to a superintelligence is the best candidate here. Which means they would be before the creation of a superintelligence.