An additional hop to the nearby stars before starting the process would delay it by 10-50 years, which costs about 10 galaxies in expectation. This is somewhere between 1e8x and 1e14x more than the Solar System, depending on whether there is a way of using every part of the galaxy.
Mass is computation is people is value. Whether there is more than 1e8x-1e14x of diminishing returns in utility from additional galaxies after the first 4e9 galaxies is a question for aligned superphilosophers. I’m not making this call with any confidence, but I think it’s very plausible that marginal utility remains high.
I’m not seeing a tradeoff. If you speed things up by a few years, that’s also a few years earlier that local superintelligences get online at all of the stars in the reachable universe and start talking to each other at the speed of light, in particular propagating any globally applicable wisdom for the frontier of colonization, or observations made from star-sized telescopes and star-sized physics experiments, or conclusions reached by star-sized superintelligences, potentially making later hops of colonization more efficient.
So maybe launching drones to distant galaxies is not the appropriate first step in colonizing the universe, this doesn’t change the point that the Sun should still be eaten in order to take whatever step is actually more useful faster. Not eating the Sun at all doesn’t even result in producing Sun-sized value. It really does need to be quite valuable for its own sake, compared to the marginal galaxies, for leaving the Sun alone to be the better option.
An additional hop to the nearby stars before starting the process would delay it by 10-50 years, which costs about 10 galaxies in expectation. This is somewhere between 1e8x and 1e14x more than the Solar System, depending on whether there is a way of using every part of the galaxy.
Mass is computation is people is value. Whether there is more than 1e8x-1e14x of diminishing returns in utility from additional galaxies after the first 4e9 galaxies is a question for aligned superphilosophers. I’m not making this call with any confidence, but I think it’s very plausible that marginal utility remains high.
I’m not claiming that marginal utility is low, just that marginal utility is much higher for other things than speeding things up by a few years.
I’m not seeing a tradeoff. If you speed things up by a few years, that’s also a few years earlier that local superintelligences get online at all of the stars in the reachable universe and start talking to each other at the speed of light, in particular propagating any globally applicable wisdom for the frontier of colonization, or observations made from star-sized telescopes and star-sized physics experiments, or conclusions reached by star-sized superintelligences, potentially making later hops of colonization more efficient.
So maybe launching drones to distant galaxies is not the appropriate first step in colonizing the universe, this doesn’t change the point that the Sun should still be eaten in order to take whatever step is actually more useful faster. Not eating the Sun at all doesn’t even result in producing Sun-sized value. It really does need to be quite valuable for its own sake, compared to the marginal galaxies, for leaving the Sun alone to be the better option.
Computational ethics is false when applied to space colonization