> Anders Sandberg <asa@nada.kth.se> wrote:
Re: planet sized supercomputers
> The only reason I can think of would be that they don't care - you
> build computing structures out of all the available matter, and the
> energy requirements tturn out to be less than the entire stellar
> output. Might happen if people settle for my 'Zeus' model of
> megastructures (cold diamondoid planet-sized system), but I better do
> some calculations on how to maximize the information
> content/production in the entire system.
These are two different things. As J/M-Brains age, they presumably accumulate more "memory" and would need to grow their storage structures. In contrast their information production is going to require energy. So you have to come up with a model where they don't want to maximize information production or are throwing away their memories. [Perhaps they are very enlightened... :-)]
As a general comment, I've tried to make all my M-Brain stuff depend on technologies that we have or will soon have. [E.g. we could start building a M-Brain today and as it evolves it could replace Pentiums with nanocomputers.] I believe that in your original paper and some work that Robert Freitas has done, information storage goes down to the Bekenstein/Bremermann bounds. Those bounds are *quite* a bit below what I can envision actually constructing. I chose circulating photons for the M-Brain storage giving you something like 10^15 - 10^27 Terabytes. This requires energy for "regeneration", but it isn't much of the power budget. There are other memory materials you could envision -- atoms in a 3D lattice, electron spin states on those atoms, nuclear spin states, etc. but it seemed to me circulating photons require the least material and have the interesting property that they can go through each other to some degree (though I don't take advantage of that in my designs).
Re: things to do as an SI
> Hurray! Hmm, maybe we should as posthumans just rearrange the solar
> system to "improve" on astrology? :-)
I'd vote for having astrology *really* work for those people who *really* beleive in it (in the upload virtual reality). I think we may be able to take advantage of that fact for our own amusement.
Robert