Re: Upload rehearsal?

From: Anders Sandberg (nv91-asa@nada.kth.se)
Date: Fri Nov 01 1996 - 10:12:29 MST


On Tue, 29 Oct 1996, Chris Hind wrote:

> Thats ok. Be as sarcastic as you like. I _rarely_ take things personally.

Great! I hope the same holds for me.

> You don't think we will build tools powerful enough or find shortcuts
> around needing an SI to do the tasks such as smart tools like windows95
> 'wizards' where the designer would just design the properties of the
> animal/plant they liked and the 'wizard' would fill in the blanks er organs
> to make the animal work.

Here is the problem: how to program the 'wizard' to make the organs? That
is a very nontrivial challenge (trust me, I have just been at an anatomy
lecture - there are embedded controllers and brilliant tricks just about
everywhere!).

Designing an organism (say a dragon) from scratch is equivalent to having
a genie machine that you can tell "Build a working spaceship!". It
requires quite a bit of intelligence, although it might manage with
"evolutionary intelligence" if it has enough brute force computing power
(to make that dragon, it would have to do millions of years of evolution
in a reasonable span of time, down to the genetic scale!).

I'm not saying it is impossible, just quite hard. Modular lifeforms might
make things easier, but life is still very complicated business.

> I watched a show on TV recently in which the girl referred to herself as
> 'we' all the time and called her four remaining personalities the 'troops'.
> She could recieve messages from all four personalities at once or so it
> seemed. Can any cases of MP disorder share data between personalities? I've
> also heard about personalities writing to other personalities by writing
> letters to each other. Is there more than one form of this disorder?

Interesting. Yes, there are many kinds. As far as I know, they might even
have different causes and ways of functioning - the divisions could exist
on different levels.

> >I was talking about running them. Let's say I want to create a cellular
> >automaton universe (the Autoverse) with roughly the same internal size and
> >complexity as the solar system. Assume we use clever compression so we
> >only have to deal with the places where matter exists, then we have around
> >10^31 kg of virtual matter, containing around 10^23 "atoms" each - 10^54
> >bits for the whole simulation. This simulation will not run in a computer
> >smaller than a star unless we can find some very clever optimizations...
>
> Do we really need to animate all the atoms which don't currently serve any
> practical use? We can simply suspend all atoms except for the ones which
> we're using and as the virtual cities grow and size and complexity of the
> universe grows, so will the processing power.

Yes and no. I would of course add this kind of optimization (why simulate
miles of pointless rock at the planetary core at the same resolution as
the surface dwellers' brains?), but it won't solve the main problem: such
a world would be inherently computationally expensive. We cannot compress
it beyond its Kolmogoroff complexity (the shortest program that generates
it at a certain timeslice), and by definition, being an interesting and
comparably complex world, it would have a complexity comparably to our
own.

I somewhat envy the protagonists in _Permutation City_, who have
literally unlimited computing power...

Until I get my neutron star computer, I plan to make small worlds.

> >The sum of human knowledge sounds nice. But I want more.
>
> I completely agree! More is better but do we even know what more is yet?
> You're looking for quantity which is 'more' but we don't even know what
> awesome technologies are going to flow out of that great expanse of 'more'.

True. But what I'm looking for is not technology, I want knowledge (and a
certain amount of wisdom). We cannot predict what is to be learned or
developed beforehand (otherwise we would know it already), but I bet it
will be interesting and beautiful!

-----------------------------------------------------------------------
Anders Sandberg Towards Ascension!
nv91-asa@nada.kth.se http://www.nada.kth.se/~nv91-asa/main.html
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 2.1.5 : Fri Nov 01 2002 - 14:35:49 MST