Re: Dyson shells are possible

Robert J. Bradbury (
Fri, 17 Sep 1999 15:15:23 -0700 (PDT)

On Fri, 17 Sep 1999, Robin Hanson wrote:

> At 01:18 PM 9/17/1999 -0700, you wrote:
> Jugaku & Nishimura may have used Papagianis 84 to *motiviate* their
> choice of a cutoff to search for, but the data they get does not
> *assume* this cutoff. The ask directly whether more than 1% of the
> starlight from each star is coming out at IR temps, and they find no.

Agreed. However J&N, and almost all other SETI searches point their telescopes at stars. So they all *assume* an advanced ETC cannot or does not "hide" the star. I would argue strongly (because of the exponential growth discussed elsewhere) that *all* long-lived advanced ETC do "hide" the star. The place you have to start is with "bright" infrared sources, not "visible" stars. I tried pointing this out to Stuart Bowyer (Acting President of IAU Cmsn 51) at the Bioastronomy conference but he was pretty closed minded about it. In contrast if you look at the Russian approach for the last ~20 years, they have consistently moved in this direction.

> >a) It is possible that at the time we are at they are already harvesting
> > 90% of the photons.
> This doesn't answer the question about the 10% we see.

Because they are at the point of "diminishing" returns. You and I don't eat all of the food on our plate if we are sufficiently "full".

> >b) H & He are not particularly good materials for building things
> > (Solid H has the hardness of butter). It is possible that "stars"
> > are the most efficient ways to convert H & He into C, O & Al, W, etc.
> This is irrelevant regarding the lost photons. Even if you don't want
> to build with star material, and even if you want stars to make metals,
> why let all those photons go?

Because you don't have the metal to harvest them! You have to allocate your materials to (a) Energy harvesting, (b) Computronium; (c) Communications. In our solar system, we do not have enough material (even after all the planets, comets, asteroids, etc.) to construct radiators that can radiate near the background temperature of the universe. We (in our solar system) would probably end up 10-40K above the background temp. That is, thermodynamically & computationally, not the most efficient place to be. The only solutions are to harvest material remotely and ship it back and that is very expensive or breed it locally (from energy) and that takes a long time.

We still see stars because there isn't enough metal in the Galaxy to "optimally" hide them all yet.

> You seem to be saying that you *know* that *all* advanced aliens
> want *only* "computational throughput", *and* that colonization
> cannot typically offer such throughput well. I'm skeptical about
> both claims. Almost all creatures today do not value only
> computational throughput; why should we expect such creatures to
> dominate the future?

I am talking about technological civilizations advanced to the limits allowed by physics. Maybe some civilizations stop short of those limits, but I would argue that is a risky position to be in, since galactic accidents in the long run will eliminate those civilizations that do not prevent or avoid those accidents. Natural selection favors those civilizations/entities that can predict & avoid or stop those things which can destroy them. Computational thoughput provides these abilities.

Humans dominate the planet because their computational ability exceeds that of almost all the other species. Say we choose to remain as "humans" and go "colonize". What happens when we encounter a Dyson Shell Supercivilization? We lose! It can see us coming and be prepared to give us a planet of little interest to it, or wipe us out of the sky. Men and ants don't compete with each other. One lives off the crumbs left behind by the other and tries to avoid getting squashed.

> And I'm sure we could ifeden.computational problems that are so
> hard that one could compute them more quickly by sending out probes
> to turn the universe into computers, rather than just using one system
> to compute with. How can you know that advanced creatures aren't
> interested in such problems?

Yes, but there is probably only a small set of computational problems where the data can be separated into logical subdivisions that do not require significant communication of inputs and outputs. Within a Dyson shell you will have a huge bandwidth and memory. Between Dyson shells you are going to be sharply constrained. It is going to be much worse than the communications difference over a fiber cable and a telegraph wire.

You win much more by figuring out the optimal computer architecture and building it locally, than you do by colonizing the nearest stars. Paraphrasing Feynman -- "There is *more* room at the bottom!". Getting smaller gains you much more than getting bigger.

Ultimately, I think Alien Dyson Shell Civilizations have to deal with the problem of dismantling their star and putting it effectively on the "outside" of their computronium. Anders has (or is) submitting a paper on some various architectures to the J. of Transhumanism. I'll be interested in seeing whether he addresses this problem.

> Even slow colonization is a problem, given how old the universe is.
> And are we to conclude that all the stars we see are zoos containing
> life? And all aliens everywhere have the same taste in zoos?

Quite possibly. Aliens could feel nostalgic for primitive life forms. Or they could also be experimenting with us. Advanced civilizations have the ability to *see* literally everything in the Galaxy and model the future paths of development extensively. They may have decided on a focus of interest and constructed an optimal computing architecture for that focus. They may have no way of knowing however whether they have computed or discovered everything and because random mutation and natural selection increases diversity and may invent something new. It may be to their advantage to maintain zoos. They may not have the same taste, but they may all have them. In that respect, Papagiannis may have gotten something right. [See "Regional Jurisdiction in Our Galaxy", Bioastronomy - The Next Steps, p. 281].

Just a side note, I don't disagree that there may be a "colonization" phase during the evolution of a galaxy. But I do think that most of ours is developed already, like our planet Earth. We do not colonize now, in fact in many cases we "uncolonized". For the most part we develop what we've got. Now, this perspective is good because it generates a testable hypothesis. As we look back in time (at increasingly distant/older galaxies), they should exhibit characteristics of less and less colonization, in theory they should become more "star-filled" and visibly brighter. Unfortunately I don't believe we have the telescope resolution to determine this yet.