Re: CRYONICS: Limits to storage & JUPITER BRAINS [was: Where to look for the answers...]

From: Anders Sandberg (asa@nada.kth.se)
Date: Sat Apr 29 2000 - 02:59:11 MDT


"Michael LaTorra" <mike99@lascruces.com> writes:

> JUPITER BRAINS
> Mike:
> Regarding a Tiplerian resurrection/emulation, Anders Sandberg has pointed
> out that, short of Tipler's Omega Point, a similar resurrection/emulation
> could be possible via AI superobjects (Jupiter/neutron star size/density AI
> brains) albeit for a less than universal coverage of the deceased. (How do
> we get on the resurrection/emulation waiting list? Maybe this Extropians
> list is IT!)At this point in time, I'd rate Anders' idea as the more likely
> of the two.
>
> Mitch:
> Being a fav subject of mine, I would also suggest that the Jupiter Brain
> emmulation concept, conjectured by Roboticist, Hans Moravec is also in the
> works, as it utilizes
> a neutron star as the the source of computation. I believe the bit-stream
> for
> such a emmulation of the entire earth history is well bellow the estimated
> 10^60 bits per second that a small neutron star is supposed to be capable
> of.

A rough estimate of what we might need: to simulate the entire
biosphere of the Earth, I once arrived at an estimate of around
10^24-10^26 bits updated at around 10^27-10^29 bits/s (some parts like
brains emulated more closely, others like the core just using simple
approximations - this is no emulation). Emulating 10^9 years of
history would take 10^30-10^32 bits, well in the range of what can be
housed in a jupiter brain, but the maximum speed is limited to around
the 10^-18 s level if we use molecular electronics and 10^-24 s if we
use nuclear interactions, which means that running through it all will
take 10^12-10^14 s / 10^6-10^8 s (I ignore parallellisation issues
here - this is most definitely wrong, but might give some useful
hints). In short, we are talking about 10^4-10^6 years/ weeks months
for simulating one timeline. Of course, we might settle for just the
tail million years and shorten the times by a factor of
thousand. Still, the amount of sampling of possible timelines will be
rather small unless you have a very big jupiter brain or neuron star.

Just some BOTE entertainment on saturday morning...

-- 
-----------------------------------------------------------------------
Anders Sandberg                                      Towards Ascension!
asa@nada.kth.se                            http://www.nada.kth.se/~asa/
GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y



This archive was generated by hypermail 2b29 : Thu Jul 27 2000 - 14:09:57 MDT