Bay 12 Games Forum

Quote from: a1s on February 07, 2009, 06:49:07 pm

Quote from: mainiac on February 07, 2009, 06:15:21 pm

I never said grey goo was easy.  I was just saying the grey goo was a prequisite for the AI making a hidden moonbase that engage in large scale mining

why?
as far as I understand your argument, it goes:
a)if AI can build moonbases then AI must have homeostasis.
b)if AI can turn all matter into nanomahines then AI must have homeostasis.
c)therefore, if AI can build moonbases then AI can turn all matter into nanomachines.
but that is logically incorrect.

However I suspect that I have failed to notice some critical part of your argument. can you fill it in?

I'm not trying to argue B.  For one thing, B is patently wrong.  Nanotech is possible without homeostasis.  Nanomachines already exist and they are not homeostatic.

err... not to be anal, but that's not what B says (the predicate is that this "goo" scenario happens, and ALL matter is turned into nanomachines of some kind). But even if we exclude B, the argument "If AI can build moonbases then AI must have homeostasis. Therefore, if AI can build moonbases then AI can turn all matter into nanomachines." seems to make no more sense. There's obviously a step missing here, and not having your knowledge of thermodynamics (except what I learned from Pixar) I can't fill it in.

Now, my reach is global, my tower secure, my cause is noble, my power is pure.

yes, you also can make computers survive aquatic
conditions,
see the strings that control the system and do anything with no assistance

first off, a short definition from wikipedia, for those who don't know what we are discussing

What are you talking about?  We've never come close to devising a completely self maintaining mechanical system. <...>
"Why, once a technological marvel is possible, does it not automatically become economically viable?"

you are confusing 2 concepts: it's one thing what we (driven by our economic greed, and limited by our current level of knowledge and know-hows) have put in place. it's entirely another what a system of machines is theoretically capable of (after all, the premise is that we have a working AI, which in the real world we don't).

In fact I'm pretty sure we (now!) can produce simple homeostatic systems in a lab. The simplest one would be just 2 magnets, one of which is stuck to the table- once you deform the system, it come back into it's original state.
Jokes aside though, we can make a machine that will make copies of itself until it's broken or out of source material, I'm fairly certain we can figure out a way to diagnose faulty ones, and we could probably have it also make another machine that will bring it source material (piled in the corner of the lab, for simplicity) and diagnose those secondary "mining" ones. So... technically if we made 2 of the making machines, had a new one produced each time an old one fails to meet diagnostic criteria, and make a new "mining" machine each time we don't get a new shipment of material, we would have a perfectly working homeostatic system, that is not living (nor economically viable, but that is already a question of time and effort spent adapting it, not actual possibility)

To put it simply, hiding a lunar base 1 mile under the surface of the moon would be friggin' impossible, let alone five.  And digging up chunks of the moon would be prohibitively expensive without humans around to run the machines.  Any equipment the AI needs on the moon has to either 1) be secretly launched from earth with no one realizing or 2) very basic use of in situ materials, keeping in mind that processing materials requires shipping up machines that will break and can't be repaired.  If the AI didn't have these constraints, then it would already have achieved grey goo status and would have destroyed humanity long before.

Any ai less advanced then the grey goo stage simply can not sustain activity without the input of human economic activity.  Humans are life and maintain homeostasis.  Machines are not life and do not maintain homeostasis.  Without homeostasis somewhere in the process, a task such as digging can only be done until the spare equipment is exhausted.  And that would come very quickly.

So the AI lunar base has to be on the surface of the moon or hidden in a crater.  Digging underground is economically impossible.

that's a pretty specieist statement. Machines maintain homeostasis by replacing any broken parts of the system, including entire machines, if need be. It is easy to observe even now, when machine systems are tended by ideologically alien (to them) engineers, and when the machines will become sentient, the quality of that will only increase.
You might say (and I will agree with you) that replacing broken parts requires a large manufacturing process (and likely some mining), however the same is true of human colonies. There is really nothing preventing the AI from setting up a self-sufficient base on the moon¹. After all we are assuming that the AI has enough resourcefulness to secretly build a small base on the moon, which means both building machines and materials- replace half of those with robot assembling machines, and necessary materials to start the process.

---

¹)well... actually there is: the moon might simply be lacking in the necessary minerals.

no, they aren't, you could be a hermaphrodite named John (but not actually "Jonathan S. Fox").

But actually you are both male and gay. or at least you were back in 2003...

so... would anyone like to GM another one of these? I kind of want to play myself...

applying for round 8, as a pararapher.
also I'd like a perma-spot, please.

Drugwars is actually one of those "classic" games, from the early 80s, like tetris, with a thousand remakes, and a surprisingly dissapointing gameplay (for their overhyped status).

That being said, I  would very much  like to see what come out of this game.