Author
|
Topic: The New Alliance Ships (Read 21236 times)
|
JonoPorter
Enlightened
Offline
Gender:
Posts: 656
Don't mess with the US.
|
but cant chang the weapons just the speed, accel, rate of fire, recharge, and aspects of the weapon but not the weapons or special itself.
|
|
|
Logged
|
|
|
|
|
JonoPorter
Enlightened
Offline
Gender:
Posts: 656
Don't mess with the US.
|
true but it has to be done in a way that it does not make the older ship obsolite or you get the DBZ effect of stuff becoming so powerofull that a stray shot wipes out a solar sytem.
|
|
« Last Edit: February 25, 2003, 12:37:26 am by BioSlayer »
|
Logged
|
|
|
|
|
JonoPorter
Enlightened
Offline
Gender:
Posts: 656
Don't mess with the US.
|
DBZ is a japanesse anime that has super heros that get a hundred time more powerfull each time they need to deafeat the newest enemy
first you have to have poeple devise a lot of new weapons and specials then make the editor to balance them. it is so you can have a custom ship to play.
|
|
|
Logged
|
|
|
|
|
JonoPorter
Enlightened
Offline
Gender:
Posts: 656
Don't mess with the US.
|
I did and i still think its possible but hard
|
|
|
Logged
|
|
|
|
|
|
Lukipela
Enlightened
Offline
Gender:
Posts: 3620
The Ancient One
|
I did, and yeah, they did But (and seeing as I'm not a programmer this might be complete and utter c**p but still...), the weapons in Qauke were always the same, so there weren't really that much of a tactical situation, was it? for humans, it'd be hiding, running and fighting, trying to get to places where you'd have the advantage while fighting. the bots would learn to kill you, dodge your tricks, but they'd always just be adapting, not really being innovative like humans can be.. Or am I completely wrong?
|
|
|
Logged
|
What's up doc?
|
|
|
|
Lukipela
Enlightened
Offline
Gender:
Posts: 3620
The Ancient One
|
Well, I rather like that idea, but no, that's not what I'm aiming at. I'm saying that with the resources we have today, at this very momnet, I think it'd be extremely hard, if not impossible to do so. In a fwe years, maybbe not, in 20, most probably. But right now? that is the question that concerns us, sin't it
|
|
|
Logged
|
What's up doc?
|
|
|
Death 999
Global Moderator
Enlightened
Offline
Gender:
Posts: 3874
We did. You did. Yes we can. No.
|
We might be able to implement a genetic algorithm that would take the tactical decisions off our hands. The fitness criteria would be pretty obvious. The problem would be designing a shape for the genetic code.
What battlefield information is available to an AI? We could develop a framework within that raw data level, perhaps working off of the existing AI code, cutting it up into interchangeable parts.
Anyone know how the AI is put together?
|
|
|
Logged
|
|
|
|
|
Flewellyn
Guest
|
Here's something I can actually comment about, as I am doing some (personal) research into artificial intelligence, and do some programming as well.
Okay, an AI as a game player is quite feasible, even though, given where AI is right now, it's not going to be as much of a challenge as a human being. Given a finite set of constraints, such as the rules of the game, you can make a robotic player that can present a moderate challenge.
Using an AI to help design ships, though? You'd be asking a program to think on the same level of abstraction, with the same degree of complexity and same ability to "jump outside the box" as a human being. We just don't have the technology for that yet: current AI is on the level of bacteria, able to react to changes in the environment, incorporate new strategies, and even anticipate a little bit...in a very primitive way. (Bacteria can do all of these things, yes, though again, very primitively.) AI cannot, at present, plan ahead, think creatively, realize that its current mode of operation is getting it nowhere, have flashes of intuition, etc.
Even gameplaying AI is very limited. People point to the Deep Blue victory over Gary Kasparov in chess as a victory for AI, but it isn't. Deep Blue was actually a rather primitive AI, which did "brute force" checking of decision trees (analysing possible moves and their consequences), with some "pruning" of the tree for moves that quickly became bad ones. But nowhere near the sophistication of a human's ability to think. Deep Blue only won because of the power of its hardware, and because it was allowed to study Kasparov's games (and only his games) ahead of time, and develop reactive strategies analogous to the genetic strategies a bacterium develops against antibiotics.
Given all of this, asking an AI to work within the ship design system and anticipate problems of game balance (which is outside the system) is impossible with current technology.
For more info on all of this, I would highly[/i] recommend a reading of Douglas Hofstadter's book Godel, Escher, Bach: An Eternal Golden Braid, which goes into great detail about all of these matters and more. Plus, it's just a fun read, and goes quickly, which is surprising given its length (1700 pages or so).
|
|
|
Logged
|
|
|
|
|