[ExI] DNA India: The transhumanity timeline

[Eugen Leitl]

On Fri, Jun 21, 2013 at 09:02:44AM -0700, Keith Henson wrote:

> If you are trying to talk to a Jupiter brain, the top level coherent
> process is what you would be talking to.  At least I think that's the
> case.

I'm not sure there can be a hierarchical assembly that can
act coherently. How would you talk to a planet full of 7 GPeople?
Large top-scale processes can relate to other large top-scale
processes, for plans that can take arbitrary amounts of time
to execute.

By why would a whale talk to an amoeba? The don't share
the same time scale, they don't share the concepts. There
is really no common base for communication at all.
 
> >> If you want to do a lot of thinking, small dimensions, lots of power
> >> and plenty of cold water are needed.
> >>
> >> http://hplusmagazine.com/2012/04/12/transhumanism-and-the-human-expansion-into-space-a-conflict-with-physics/
> >
> > Too many assumptions there. "Twenty million subjective years to get an answer back is just ridiculous." --
> > you're thinking like a human here.
> 
> I agree with you.  Unfortunately I don't know how to think other than
> like a human.  Any advice?

I would just look at which behaviour constraints are available
for systems of all scales of complexity, including very small 
nonsentient systems and dauer spores, up to very large scale
systems and supersystems in the context of Darwinian evolution. 
It seems that the only constraints there are that of metabolism 
and relativistic travel. Motivation differs across scales, and
we definitely can't assume any human motivators apply.

We already have thousands of volunteers applying for a one-way
essential suicide mission. Extreme environments like Mt. Everest
have plenty of dead bodies to be used as landmarks. Even people
do things which they apparently shouldn't be doing.
 
> > "Due to this line of thinking, I no longer think it?s practical to surround a star with computronium." --
> > why, obviously Amerigo Vespucci and Cristobal Colon never set sail, so the colonization never happened,
> > so Keith Henson never happened. Since you're obviously there, there seems to be a problem with your
> > line of reasoning. It obviously has to do with the need to communicate with everyone else within
> > a given minimal time, which is a bogus requirement. Back then, the opposite site of the Earth
> > was as remote as the next star in terms of communication latency. Yet it still got settled.
> 
> By the time of Magellan, it only took 3 years sail around the Earth.

Yes, 3 years is basically interstellar distances. I can now send
ping any node on the Internet in less than a second. They didn't
have telecommunications back then, messages took many years hitching
rides on ships and travelers, and yet people expanded in increments
just fine. People just sailed into the seas without having a good
story that they will ever return, and countless perished. Yet we're
all descendants from these survivors, and the wanderlust is in
our DNA. We don't contract, we expand. We would have expanded into
space a long time ago, if we could.

> Point accepted that the Earth got settled, or rather conquered since
> the entire planet had been settled long before Europeans started

Yes, precisely, the planet was already settled by people with
much more primitive tools and even no concept of global communication
at all -- the language bareer alone would make such a thing nonsensical.

> exploring.  But that's not the point.  Is the best substrate for
> thinking beings a fog of computronium around a star?  I don't think so

We know that computronium is the best substrate for computation,
by definition. Everything else is about how much power you can
dissipate. You'll need enough 4 K background visible to dump
your heat into, and you'll expand into utilizing all available
energy flux, so it means that you have a halo of nodes around
the star. Whether the nodes are thin, or fat is not really that
important, other that they've spread across space lighthours
across, and other such sphere are across other stars lightyears
away, and so on.

> for engineering reasons of waste heat and latency.  It's a question
> akin to where people live today, i.e., a desert supports only a small
> fraction of those in lush surroundings.  For the uploaded who want to

Lush surroundings = lots of solar flux, and atoms to make use of it.
Deserts = interstellar space with no flux, and even no fusion fuel.

> think fast, plenty of power, lots of cold water and a compact
> community are what you need.  By comparison, a disbursed cloud of
> computronium would be like a bleak desert.

No, because the cloud is immaterial, just what's in the node
is important. Whether a node is cm^3, m^3 or even larger, it's
not important. The point is that that there are many nodes like
that (about Avogadro number of human equivalents in just this
solar system alone). They're not all in one location, because 
you couldn't compute that way without suffering the corium or
plasma cloud as failure modes. I don't think you can compute
warmer than 700 K, and arguably it could be useful to go lower,
perhaps a lot lower.
 
> > ?Just to give you a sense of what microseconds are, it takes you 500,000 microseconds just to click a
> > mouse. But if you?re a Wall Street algorithm and you?re fivemicroseconds behind, you?re a loser.? --
> > the problem with HFT is not that is fast (it is), but that it's systematically fraudulent.
> 
> No argument there.  Have you noted that HF trading is way down?  Seems
> the competition took the profit out of it and now there is hardly
> enough to pay the electric bills.