[General Battuta]

Uh, I know, I said that.

[Aardwolf]

I don't have a print/web source, but I have a weapons tech nut and my dad, a molecular physicist.

[General Battuta]

Hey, cool. I would like to learn more about this hull-shattering X-ray effect. I have a good amount of physics knowledge, but I hadn't imagined that EM radiation at that wavelength would shatter much of anything.

[Retsof]

I suppose that high-energy radiation would do a good job of knocking atoms around.  And what is your hull made of? (Rhetorical question)

[Zoltan]

In the near future? Extensive use of high-velocity kinetic attacks and shotgun-type weapons. No railguns, because it's doubtful they'll be perfected to a level where they're practical to lift into orbit and power with fuel or solar cells.

Relatively ineffective or simply nonexistent armor -- these ships are going to be built in Earth orbit or launched from the Earth's surface, and armor is heavy. As someone said, compartmentalization is the order of the day.

No human crews.

A focus on destroying not enemy ships but enemy communications and surveillance satellites.

Nukes, though without EMP or blast wave they'll be less effective.

Did I mention the lack of human crews? They'd be remote-operated, or autonomous.

Lastly, fuel limits. Maneuvering and orbital dynamics would be limited by chemical fuels.

Even more speculatively, I think that swarms of tiny satellites might be cheaper and more popular than larger vessels. Why risk something as expensive as a combat Shuttle if you can instead deploy a flock of tiny, stealthy, nuke-bearing nanosats? You could even use Shuttles (or their near-future successors) as carriers.

Oh, and orbital attacks on launch sites. I bet space wars would be very short.

I think that human crews would be critical for space combat, or at least it would need to be controlled at short range, which makes the idea of remote control pointless anyways. While communications delay is negligible on Earth, it would be too long for operating anything at great distance.

[Aardwolf]

Well yes, you need someone to tell the ships what to attack, whether they are onboard or not. But if you assume they are not on board the ships, then you don't need crew compartments, as much hull space, etc., and the ships become more of a machine than a vehicle.

Basically, then, as long as you can communicate enough to tell them what your objectives are, and keep an occasionally updated algorithm for completing such objectives, you can just deploy them and have them do your bidding. Your ships can communicate among themselves faster and better than if you were to rely on voice-broadcast, and thus could coordinate attacks/defenses/strategies/tactics. Although some override for strategies and campaign objectives might be necessary, tactical command would be an unnecessary, and in some cases stupid tool. Computers could (if we knew how) be made much more effective at tactical/strategic thinking than humans.

Of course, it would be stupid to put the computers in charge of picking what an enemy is; if they detect hostility, they should merely defend themselves and make every effort to report back to human government/command for confirmation that they may engage.

[Zoltan]

Well yes, you need someone to tell the ships what to attack, whether they are onboard or not. But if you assume they are not on board the ships, then you don't need crew compartments, as much hull space, etc., and the ships become more of a machine than a vehicle.

Basically, then, as long as you can communicate enough to tell them what your objectives are, and keep an occasionally updated algorithm for completing such objectives, you can just deploy them and have them do your bidding. Your ships can communicate among themselves faster and better than if you were to rely on voice-broadcast, and thus could coordinate attacks/defenses/strategies/tactics. Although some override for strategies and campaign objectives might be necessary, tactical command would be an unnecessary, and in some cases stupid tool. Computers could (if we knew how) be made much more effective at tactical/strategic thinking than humans.

Of course, it would be stupid to put the computers in charge of picking what an enemy is; if they detect hostility, they should merely defend themselves and make every effort to report back to human government/command for confirmation that they may engage.

I agree that using computers would be the best way, but it would basically boil down to whatever ship has the best computer will win. It would be like playing chess with a computer... I know that everyone wants the best weapons possible, but where is the fun in being invincible.

[Aardwolf]

The fun isn't in being invincible, it's in what you can do while you're alive that you can't do when you're dead. Or in pain, for that matter.

And it wouldn't just be about the best computers, that's like saying there's one best strategy or unit in a strategy game.

[Zoltan]

The fun isn't in being invincible, it's in what you can do while you're alive that you can't do when you're dead. Or in pain, for that matter.

And it wouldn't just be about the best computers, that's like saying there's one best strategy or unit in a strategy game.

That is something I totally disagree with. In a strategy game you are supposed to be able to win, everything is balanced, in war, you are not supposed to have a chance, and there is no balance. One more thing, have you ever played chess against a computer?

[Man-Whore]

in war, you are not supposed to have a chance, and there is no balance.

"All is fair in love and war."

Okay, that's bull****. But it's a famous quote. . .

[General Battuta]

Spacecraft will be largely controlled by computers anyway, even with human crews. It's not a huge difference in terms of actual combat performance.

[NGTM-1R]

And it wouldn't just be about the best computers, that's like saying there's one best strategy or unit in a strategy game.

It might well come down to that, just as frankly there can be such other things. The Berserker series alludes to it more than once; in war against a race of somebody's ultimate weapon planetkilling AIs gone rogue, it is, when you get down to it, fought between our computers and them more than it is between us and them. Because the computers can handle it better than we can.

[Aardwolf]

I understand that, what I meant was there are other factors, like starting conditions, differences in technology, etc. that all come into play.

[ChronoReverse]

What about small drones that are designed to dodge and then collide with their targets?  Basically larger missiles with intelligence.  As someone pointed out, it's not too hard to hit missiles that are flying straight in, but surely anti-missile coverage on a frigate would have some weak points that would, with enough weaving drones, allow a hit.

In any case, 100 kilometers would be "point blank" in a proper space battle.  If modern artillery has self-correcting projectiles, surely the future would too.

[Polpolion]

And it wouldn't just be about the best computers, that's like saying there's one best strategy or unit in a strategy game.

It might well come down to that, just as frankly there can be such other things. The Berserker series alludes to it more than once; in war against a race of somebody's ultimate weapon planetkilling AIs gone rogue, it is, when you get down to it, fought between our computers and them more than it is between us and them. Because the computers can handle it better than we can.

I doubt that in realistic combat anyone would make a sentient AI to do the fighting for them. I could see RC drones and a maybe a few proper, independent drones, but nothing like that.

[Aardwolf]

The only reason people wouldn't do it is because of all the sci-fi stories where it goes wrong, which have ridiculous premises like that the AI could choose that the humans are its enemy, etc.

[Polpolion]

What would be the benefit to having a sentient AI as opposed to a non-sentient AI of similar capabilities?

[Aardwolf]

I think a non-sentient AI would be superior, actually... (but I can't tell who that was addressed at)

[General Battuta]

The sentient AI would be more flexible.

Nonsentient computers (like the ones we have today) are constantly making stupid mistakes. They're very poor at common-sense, even if they're great at complex analysis or processes.

[Nuke]

The sentient AI would be more flexible.

Nonsentient computers (like the ones we have today) are constantly carrying out the stupid mistakes of the programmer. They're very poor at common-sense, even if they're great at complex analysis or processes.

fixed