Eliezer S. Yudkowsky wrote:
> Billy Brown wrote:
> > and one mega-nuke per comet is a cheap
> > trade - I can easily take out as many comets as you care to throw.
> I'm curious. What keeps me from throwing a mega-nuke at the planet?
> Why can't I use that to defend the comet?
Ah. Now comes the interesting stuff.
Nothing keeps you from using a mega-nuke on the planet, of course. Good missile defenses can make it a challenge, but a determined attacker is bound to land one eventually (and this ignores the terrorism problem). That's why you either need to move into space, or live very deep underground.
You can put anti-missile systems on the comet, but its hard to protect - one hit will take it out, and a near miss will take out your defenses. Add energy weapons to the equation, and the comet becomes pointless - you're better off just building a fleet.
> Also, how would you defend against a laser the size of a small moon, or
> even a system of mirrors focusing 50x normal solar intensity onto
> Earth's surface? As near as I can tell, the only defense is to prevent
> it from being built.
The mirror system is no threat - we just tell everyone to stay inside their diamondoid homes while we blow it up. Besides, you can hardly build it in secret. :-)
Moon-sized weapons are a different ball game - they imply much more advanced technology. IMO, you can't build such things unless you have SIs doing the engineering, in which case who knows what they'll come up with.
If humans are still in charge, then all the mass in the solar system is likely to be claimed well before such things can be built - which means you convert the planets into overgrown space fortresses. We don't have any energy sources handy that could blow up something that big, so aggressors would probably be reduced to staging physical invasions - bombard the surface with mega-nukes, then send in a few billion war bots to try and dig the defenders out.
> > If both sides have the same tech base, there is no obvious way to tell
> > whether attackers or defenders will have the advantage..
> As far as I can tell, the largest mass wins - BUT:
Having more mass in a nano-war gives the same kind of advantage as having a bigger economy does in a conventional war. That is usually decisive, but not always. The Eastern Front in WWII is a good example of this - the Germans were routinely outnumbered 10:1 or more, with even more lopsided disparities in heavy equipment, yet they came within spiting distance of winning.
> 1) The defender has to keep a living, fragile populace safe.
In the early phase, yes. By the time you start tossing moons around, expect virtually everyone to be uploaded, cyborged, distributed, or otherwise non-fragile. Of course, you still need to have a secure rear area - but it could easily be deep underground.
> 2) The attacker has a whole solar system to hide in.
You can't hide an economy. Individual units could easily hide in the outer system, but large-scale nanofacturing will produce huge amounts of waste heat that would be very visible even at large distances. Besides, combining future sensors with even primitive AI would let us build a system capable of tracking every drive flame and macro-scale spacecraft in the system, which pretty much rules out hiding large-scale movement.
> 3) You'd better control the entire mass of the solar system, especially
> Earth's core, if you don't want someone to take it away from you.
Absolutely. You'd also better do constant training and spend heavily on R&D, because yesterday's defenses are useless tomorrow.
Billy Brown, MCSE+I