[ExI] simulation as an improvement over reality

[Adrian Tymes]

On Tue, Dec 28, 2010 at 12:04 PM, Eugen Leitl <eugen at leitl.org> wrote:
> Very few people do real things in space. NASA is one of them.

So, now, is SpaceX.  I've got more faith in the potential for private
foundations to develop space than I do in NASA, despite NASA's
long history of space experience.

>> RepRap project, by means of comparison, has at least a little
>> funding.)
>
> You keep blowing my mind with these off-hand remarks of yours.

I keep thinking everyone here knows about these sorts of things.

http://reprap.org/wiki/Main_Page
http://blog.reprap.org/2010/01/build-better-reprap-80000-prize.html

I'm concerned that most on RepRap have been overly celebrating
its ability to create a raw structure, and ignoring the fact that it
can't create critical tool bits - electronic circuits, print heads, and
so forth.  From reading up on the terms of that prize, it seems I'm
not the only one.

Just to toss this out for discussion (and, if anyone responds on
this, please change the subject line): what is the simplest set of
tools - "simplest" as in "easiest for humans with presently
available tools and technology" - that can be used to manufacture
itself, excluding the null set and uncontrolled biological entities?
(Null set excluded for obvious reasons.  The biological exclusion
applies to, e.g., other humans - who can be manufactured by two
unskilled laborers, but are by no means designed or controllable
in the way a hammer is designed and controllable.)

> If you don't like NASA, just wait another decade or two.
> The budget gutting will continue until morale improves
> (or the patient expires).

I don't think it'll be entirely defunded any time soon.  It's too useful
as a distraction for certain types - think "bread and circuses".

> And the first guys in control
> of the Moon own the entire solar system.

Depends on what they do with the Moon.  It is certainly possible
to have an unbeatable winning strategy from that point, much
like how in chess, you can get into a position where victory is
guaranteed if you play perfectly.  But frankly, I suspect that if
any national government gets into that position within the next
30 years, they'll be playing as a rank amateur - perhaps waiting
around, perhaps focusing on national pride instead of building up
an infrastructure.

Actually, a more relevant metaphor occurs to me, for anyone
who's familiar with StarCraft (or real time strategy games in
general).  Take one player, as Terrans, who pushes up the tech
chain and builds a battlecruiser.  Said player cruises around,
flaunting it while taunting - not getting close enough to engage,
just trying to intimidate the other players into surrendering.
Then someone zerg rushes said first player.  (The winning
strategy would have been, "build more battlecruisers, keeping
them at home for defense until it's time to attack".)

> Semi-soft landing of >100 kg parcels on Luna affordably is the
> most expensive and critical aspect of the venture. Scaling down
> robotics and moving towards self-replication closure is a much
> easier problem in comparison. One you could attempt at leasure
> in groundside labs, and lunar simulators. If people do it right
> it will be competitive, multi-team, and prize-based.

I disagree that robotics is much easier.  Again, see RepRap and
the problems they've encountered.  Semi-soft landing, one can
attempt on Earth - indeed, I've seen setups that were testing that
very thing (albeit with under 100 kg, though the full-up systems
can certainly be tested once smaller replicas prove out the
concept).

However, I do agree that getting stuff to Luna is more expensive.
The majority of that, in turn, is simply the cost of launching
stuff off of Earth.  Intense R&D is being done today - again, see
SpaceX.  This is "easier" in that one simply needs to throw
enough money at the people already working on this.