The increase in the number of Everett branches could be very important.
Let’s suppose the future unfolds in a positive scenario for us: someone creates an aligned AI, it takes over the world, and colonizes the light cone.
Let’s suppose the Everett interpretation of quantum mechanics is correct.
Let’s suppose we care about happy lives regardless of whether there are their copies (imagine that tomorrow you meet your identical twin who has experienced events and feelings almost identical to yours. Would that mean your life was unimportant, or that together your lives were just as valuable as the life of a single unique person?)
Let’s suppose there’s no something in future that destroy all values like tortures or baby eating aliens.
If you agree with all these premises, then our civilization would need to use the energy of stars to further accelerate the branching of the universe. Even a single new split would mean that, across all universes, there are twice as many happy people in total.
There aren’t discrete “splits” in quantum mechanics, and consequently no way to accelerate them. Branches are a lies-to-children simplification that you can employ in a few toy scenarios, but fail to be useful in pretty much everything else.
This proposed utility function term is a gazillion times easier to satisfy for each second earlier that you start; therefore, you can take a dollar to the casino and double it until you lose, and if you got enough winnings buy all the AI labs and implement your vision. After you tried that, any further contributions would be a rounding error so you might as well assume away that term in the utility function.
I would agree with you if I was really sure about all the chains of the argument, but I can’t never be really sure because of epistemic humility, so I wouldn’t do a such of weird plan.
Fine, I’ll go buy a damn lottery ticket then.
If you care about almost-identical copies, why not care about identical copies? Then there’s no need to split :)