A niche for space fighters

SF: discuss futuristic sci-fi series, ideas, and crossovers.

Moderator: NecronLord

User avatar
Stark
Emperor's Hand
Posts: 36169
Joined: 2002-07-03 09:56pm
Location: Brisbane, Australia

Post by Stark »

frogcurry wrote:In answer to Starks last comment, I don't think the person who started the topic is concerned about going for that level of depth of realism. Not when they are considering empires and pirates in space in the first post...
So why the hell is it even an issue?

This sort of thing irritates me, because plenty of people want to CLAIM 'hard' scifi, or 'harder than x' scifi, because they think it makes them 'better'. If you've got standard scifi shit, why bother trying to ruin your story by pretending to be all realistic? Write the story you want to write, for fucks sake. :roll:
User avatar
Teleros
Jedi Council Member
Posts: 1544
Joined: 2006-03-31 02:11pm
Location: Ultra Prime, Klovia
Contact:

Post by Teleros »

FTL needn't be a staple of sci-fi stories, but it is certainly a regular feature in it.
You COULD have a sci-fi in which the controlling mind of the craft in FTL needs to have the ability to reason in a non-machine/computer manner to successfully transition from one place to another.
Give it time and we'll have sentient computers before you know it.
No, it doesn't. Not enough to make a difference. It's not free to attach weapons to a scout. Weapons have mass. You have to pay for that mass in delta-V, other equipment, or money. There's no point doing that for the chance to shoot down an enemy scout once every billion sorties or whatever ridiculous number it turns out to be.
1. Nonsense. If for the sake of argument we decide to investigate a cubic AU with Earth at its centre and dispatch a scout or two there, with no horizon in space there's a good chance opposing scouts will detect each other. Now if one is close enough to get a missile off and take out the opposing scout, no more data. Now you might be able to get the necessary data from a capital ship's long range telescopes, but (i) you can't expect that in every single situation and (ii) the closer you get the better resolution you're going to get - so there's a potential need to have scouts of some sort nearer whatever it is you're scanning.
2. Of course it's not free to attach weapons, but if the data scouts bring back is that important, a navy might want to stick a weapon on it after all to deny the enemy the same data, or to protect against such an attack. The result may be a drone rather than a manned fighter, but if we assume that the data these scouts can get is important enough to try and deny it to the enemy, it's important enough to defend too, so you'll end up with an armed recon drone at the very least.
Why not put him in a cutter, since you need one anyway?
My thinking was if you have say 20 people per cutter, they're concentrated in one place, whereas with fighters, they can be in a lot more places at once. I suppose the best analogy would be a police van vs several patrol cars: the patrol cars can cover a lot more ground and you do have a policeman on the scene, which it seems is generally preferred over CCTV (ie the drone reporting back to HQ).
Also remember that not all police work involves boarding smugglers and the like. What if we've got a couple of merchant ships squabbling over some minor / perceived infraction? I'd say it would be a waste sending over the cutter (which could be doing the anti-smuggling bit) and that a human would carry greater weight psychologically than a voice / talking head coming from our drone. What if local law requires officers to personally check the credentials of incoming ships? Now probably they should change said local law, but my point is that there are some things where a human is better suited than a machine, but where a cutter is OTT.
As to whether it justifies the expense, well that would probably depend on things like quantity & value of traffic, not to mention training and manufacturing costs, which is up to whoever creates the universe.
First, there's no reason to have them, or a carrier. Second, there are two damn good reasons not to throw them into battle "just because": the cost of the spacecraft and the lives of the pilots. Real navies don't throw auxiliary craft away in battle just because, so why would a space navy do it?
1. I never meant for the scenario to be a regular feature in space combat. Consider my earlier armed scouts being transported in a convoy when they're attacked. As for the carrier, well they've got to launch from something unless they can keep up with an FTL fleet on their own power & supplies (!), even if its from a proper capital ship (maybe "mothership" rather can "carrier" would have been a better term here).
2. Might it not use everything it had if it desperately needed every possible edge? Under any normal circumstances I'd agree, but you know what they say about desperate times. In any other scenario I'd have the commander of that carrier strung up for sheer incompetence for wasting lives and money.
3. Yes all right I'm making assumptions that there are space fighters in the first place - my point here was to give a possible use for them in a fleet engagement, although as I hope you can see I think we're in agreement that they've no place in a battle between the big ships, at least outside of exceptional circumstances (eg sheer desperation) or some in-universe reason (knights in space and all that).
User avatar
SWPIGWANG
Jedi Council Member
Posts: 1693
Joined: 2002-09-24 05:00pm
Location: Commence Primary Ignorance

Post by SWPIGWANG »

I think we are having a failure of imagination here.

It is obvious that a 30ton vehicle probably can't get to much of anywhere and fit enough equipment. However the fighter itself had hardly remained the same over time.

Sopwith Camel Weight: 660kg
F-15E weight: 36,000kg

We are looking at a size increase of 60 times in one century. In a few development cycles like this, an aircraft is going to weight quite a high number.

Make the fighter a 10,000ton combat vehicle, and the technology would actually fit and everything. As for the why there is a crew there? Well, lets just say AI quite aren't there and there is a need for a crew in the loop.
User avatar
Teleros
Jedi Council Member
Posts: 1544
Joined: 2006-03-31 02:11pm
Location: Ultra Prime, Klovia
Contact:

Post by Teleros »

SWPIGWANG wrote:I think we are having a failure of imagination here.
Have to say I've been feeling the same way. Manned fighters are nice and emotive, but not particularly realistic sadly.
Make the fighter a 10,000ton combat vehicle, and the technology would actually fit and everything. As for the why there is a crew there? Well, lets just say AI quite aren't there and there is a need for a crew in the loop.
David Weber does something like this for the LACs in the Honor Harrington universe. Anyone got figures for the mass of say Shrike LACs?
petesampras
Jedi Knight
Posts: 541
Joined: 2005-05-19 12:06pm

Post by petesampras »

SWPIGWANG wrote: Well, lets just say AI quite aren't there and there is a need for a crew in the loop.
Well, this is your problem right here. Given how far computers and software have advanced in just 60 years, it is highly unlikely we won't have sufficient A.I. by the time 'outer space combat' becomes an issue.
User avatar
Teleros
Jedi Council Member
Posts: 1544
Joined: 2006-03-31 02:11pm
Location: Ultra Prime, Klovia
Contact:

Post by Teleros »

I expect the thing limiting the computers would be things like power supplies. It might be that the AI grabs too big a share of the fighter's limited fuel for example. And we can always invoke an in-universe reason to limit computers (fear of sentient computers, ban on AI research, technology backsliding...).
User avatar
RedImperator
Roosevelt Republican
Posts: 16465
Joined: 2002-07-11 07:59pm
Location: Delaware
Contact:

Post by RedImperator »

Teleros wrote:
No, it doesn't. Not enough to make a difference. It's not free to attach weapons to a scout. Weapons have mass. You have to pay for that mass in delta-V, other equipment, or money. There's no point doing that for the chance to shoot down an enemy scout once every billion sorties or whatever ridiculous number it turns out to be.
1. Nonsense. If for the sake of argument we decide to investigate a cubic AU with Earth at its centre and dispatch a scout or two there, with no horizon in space there's a good chance opposing scouts will detect each other. Now if one is close enough to get a missile off and take out the opposing scout, no more data.
Do you know how big a cubic AU actually is? Eight light minutes to a side. Even if the scouts do see each other, lightspeed lag makes accurate targeting completely impossible, and you can't carry weapons on a drone big enough to hurt anything at any appreciable range anyway. The chances of two drones being in combat range of each other are astronomical.
Now you might be able to get the necessary data from a capital ship's long range telescopes, but (i) you can't expect that in every single situation and (ii) the closer you get the better resolution you're going to get - so there's a potential need to have scouts of some sort nearer whatever it is you're scanning.
2. Of course it's not free to attach weapons, but if the data scouts bring back is that important, a navy might want to stick a weapon on it after all to deny the enemy the same data, or to protect against such an attack.
If an enemy navy is trying to shoot down drones, it won't send armed scouts, it will send hunter-killer drones or just have full sized ships blast it. A scout won't be able to defend itself against the former without compromising its mission, and it won't be able to defend itself against the latter at all. The only good defense for a scout drone is to run away as fast as it can; weapons only slow it down.

It's worth looking at real life spyplanes here--most of them are completely unarmed, as are most unmanned reconnaissance drones. The armed ones aren't equipped for air-to-air combat, either: they attack targets of opportunity on the ground. A scout in space needs speed, good instruments, and a low profile so maybe it's not noticed. Not guns.
The result may be a drone rather than a manned fighter, but if we assume that the data these scouts can get is important enough to try and deny it to the enemy, it's important enough to defend too, so you'll end up with an armed recon drone at the very least.
An armed recon drone will be just as helpless as an unarmed one, with the added disadvantage of having less delta-V.
My thinking was if you have say 20 people per cutter, they're concentrated in one place, whereas with fighters, they can be in a lot more places at once.
Twenty people doing something useful in one place is better than the same twenty people being useless in twenty different places.
I suppose the best analogy would be a police van vs several patrol cars: the patrol cars can cover a lot more ground and you do have a policeman on the scene, which it seems is generally preferred over CCTV (ie the drone reporting back to HQ).
It's a lousy analogy because space isn't anything like a terrestrial city and neither will space traffic or space law enforcement. The best analogy is the Coast Guard, and guess what: they don't send people out in one-man motorboats.
Also remember that not all police work involves boarding smugglers and the like. What if we've got a couple of merchant ships squabbling over some minor / perceived infraction? I'd say it would be a waste sending over the cutter (which could be doing the anti-smuggling bit) and that a human would carry greater weight psychologically than a voice / talking head coming from our drone. What if local law requires officers to personally check the credentials of incoming ships? Now probably they should change said local law, but my point is that there are some things where a human is better suited than a machine, but where a cutter is OTT.
As to whether it justifies the expense, well that would probably depend on things like quantity & value of traffic, not to mention training and manufacturing costs, which is up to whoever creates the universe.
How many times do you need to be told anything small enough to be reasonably considered a fighter isn't going to have any kind of range, making it useless very far away from the mothership and incapable of exceeding the range of the mothership's weapons, so you might as well just send the damn mothership? And again, if this kind of thinking worked in real life, the Coast Guard would send one guy in a motorboat out on patrol, and yet it doesn't.
1. I never meant for the scenario to be a regular feature in space combat. Consider my earlier armed scouts being transported in a convoy when they're attacked. As for the carrier, well they've got to launch from something unless they can keep up with an FTL fleet on their own power & supplies (!), even if its from a proper capital ship (maybe "mothership" rather can "carrier" would have been a better term here).
2. Might it not use everything it had if it desperately needed every possible edge? Under any normal circumstances I'd agree, but you know what they say about desperate times. In any other scenario I'd have the commander of that carrier strung up for sheer incompetence for wasting lives and money.
3. Yes all right I'm making assumptions that there are space fighters in the first place - my point here was to give a possible use for them in a fleet engagement, although as I hope you can see I think we're in agreement that they've no place in a battle between the big ships, at least outside of exceptional circumstances (eg sheer desperation) or some in-universe reason (knights in space and all that).
This isn't a "use", this is an accidental circumstance where there's nothing to lose, so you might as well chuck them out and hope maybe they'll soak up some bullets for you. You might as well claim bricks have a use as infantry weapons, on account of the fact that an infantryman who runs out of ammo in urban fighting might start throwing bricks at the enemy.
SWPIGWANG wrote:Make the fighter a 10,000ton combat vehicle, and the technology would actually fit and everything. As for the why there is a crew there? Well, lets just say AI quite aren't there and there is a need for a crew in the loop.
Someone needs to inform the Argentines that the AI on their Exocets wasn't "quite there" when they sank HMS Sheffield. This is space: everyone can see everyone else, and the math problems required for navigation are done much better by computers. Point the missile at the target and send it on its merry way and don't bother wasting money and lives putting humans on board.
Image
Any city gets what it admires, will pay for, and, ultimately, deserves…We want and deserve tin-can architecture in a tinhorn culture. And we will probably be judged not by the monuments we build but by those we have destroyed.--Ada Louise Huxtable, "Farewell to Penn Station", New York Times editorial, 30 October 1963
X-Ray Blues
User avatar
Batman
Emperor's Hand
Posts: 16333
Joined: 2002-07-09 04:51am
Location: Seriously thinking about moving to Marvel because so much of the DCEU stinks

Post by Batman »

SWPIGWANG wrote:I think we are having a failure of imagination here.
Yep. On your part.
It is obvious that a 30ton vehicle probably can't get to much of anywhere and fit enough equipment. However the fighter itself had hardly remained the same over time.
Sopwith Camel Weight: 660kg
F-15E weight: 36,000kg
We are looking at a size increase of 60 times in one century. In a few development cycles like this, an aircraft is going to weight quite a high number.
The term 'NO' comes to mind, ignoring for the moment that there ARE combat aircraft that weigh in at hundreds of tons-they're called bombers.
Ignoring the fact that the Strike Eagle ISN'T a fighter, it's a STRIKE FIGHTER. Well and the fact that unlike spacefighters, airfighters can overcome a lot of the mass penalties thanks to aerodynamic lift.
'Next time I let Superman take charge, just hit me. Real hard.'
'You're a princess from a society of immortal warriors. I'm a rich kid with issues. Lots of issues.'
'No. No dating for the Batman. It might cut into your brooding time.'
'Tactically we have multiple objectives. So we need to split into teams.'-'Dibs on the Amazon!'
'Hey, we both have a Martian's phone number on our speed dial. I think I deserve the benefit of the doubt.'
'You know, for a guy with like 50 different kinds of vision, you sure are blind.'
petesampras
Jedi Knight
Posts: 541
Joined: 2005-05-19 12:06pm

Post by petesampras »

Teleros wrote:I expect the thing limiting the computers would be things like power supplies. It might be that the AI grabs too big a share of the fighter's limited fuel for example.
Except that computers are getting more and more efficient in terms of computing power versus power requirements. I think the life support of the pilot will likely be a far bigger drain on fuel than the A.I.
And we can always invoke an in-universe reason to limit computers (fear of sentient computers, ban on AI research, technology backsliding...).
Of course, this type of thing only realistic option, in my opinion.
User avatar
Beowulf
The Patrician
Posts: 10619
Joined: 2002-07-04 01:18am
Location: 32ULV

Post by Beowulf »

Batman wrote:
SWPIGWANG wrote:I think we are having a failure of imagination here.
Yep. On your part.
It is obvious that a 30ton vehicle probably can't get to much of anywhere and fit enough equipment. However the fighter itself had hardly remained the same over time.
Sopwith Camel Weight: 660kg
F-15E weight: 36,000kg
We are looking at a size increase of 60 times in one century. In a few development cycles like this, an aircraft is going to weight quite a high number.
The term 'NO' comes to mind, ignoring for the moment that there ARE combat aircraft that weigh in at hundreds of tons-they're called bombers.
Ignoring the fact that the Strike Eagle ISN'T a fighter, it's a STRIKE FIGHTER. Well and the fact that unlike spacefighters, airfighters can overcome a lot of the mass penalties thanks to aerodynamic lift.
Define what a fighter is. BTW, a Strike Eagle isn't substationally different in airframe from a F-15C. Oh, and the F-22 has the same MTOW as the F-15E. Also note that the F-117 has negligible air to air capability (theoritically cleared for an AIM-9, but this has never been used).
"preemptive killing of cops might not be such a bad idea from a personal saftey[sic] standpoint..." --Keevan Colton
"There's a word for bias you can't see: Yours." -- William Saletan
User avatar
Nyrath
Padawan Learner
Posts: 341
Joined: 2006-01-23 04:04pm
Location: the praeternatural tower
Contact:

Post by Nyrath »

Teleros wrote:And we can always invoke an in-universe reason to limit computers (fear of sentient computers, ban on AI research, technology backsliding...).
The trouble with that kind of solution is it has to be universal. Or the one evil nation that isn't afraid of sentient computers/has no ban on AI research/has technology that progresses will rule the galaxy.

In another forum, somebody suggested that all the astromilitaries of all the space faring nations used fighters instead of missile buses because if you were not risking a human pilot, how just could your cause be?

I pointed out the same flaw: it has to be universal. As an analogy, imagine a world where the soldiers of all nations did not use firearms, only swords. If you did not engage your enemy within arm's length, how just could your cause be?

Until one rude nation spoils a good time for everybody else by giving their soldiers machine guns.

After the other nations get tired of having their soldiers slaughtered at long range, while pathetically waving their swords, they will have no choice but to also adopt machine guns.

So if you are going to go this route, you have to explain how you will enforce it on everybody with absolutely no exceptions.
User avatar
SWPIGWANG
Jedi Council Member
Posts: 1693
Joined: 2002-09-24 05:00pm
Location: Commence Primary Ignorance

Post by SWPIGWANG »

I think you people are overstating the advantages of the AI.

It is not that AI are ineffective or can't do what a human can do once given proper development, but adding a human is hardly something as detrimental as using swords to charge the enemy.

Inside a 10,000ton "Mega-Fighter", something like crew life support is almost trivial compared to the mega-tons of energy being expanded in the engine and weapons and systems build to withstand it. Adding a man in the loop merely reduces the weight/cost efficiency by an insignificant fraction.

So why send a human? Well there are a few possible reasons:
1. Culture reasons: There has been beliefs of board members that war should always be fought by people or else we'd forget the meaning of it. While modern culture attempts to perserve life, a future culture with biological immortality may not have the same view or one may suffer impossible population explosion. The possible culture reasons for this sort of thing can be easily derived from the desire of having manned fighters to begin with. For example, serving could be a rite of passage ritual for a culture.

2. The specific capacities of human intelligence. While things like computing fuel efficient course trajectory and things are best done with a computer, when one is fighting an human opponent (even if a human opponent far from the front in a command center), one may want some sort of understanding of human context. People many simply not trust computers to do things like knowing when to accept surrenders, when to attack suspecious freighters, when to actually perserve enemy forces to strategic reasons (eg. 3rd fraction joining the fray) and all the subtle things.

While it would be possible to build a computer that can do this, there is simply no estimate of how hard it would be, and one can just argue that it is not achieved in the universe. In anycase, one could argue against the existence of humans at all if given that level of technology.

Now you may argue that those things should be done in a command ship not a "frontline" fighter, but lets just assume that the time-scales of decision making is short enough that a frontline unit needs to do it.
petesampras
Jedi Knight
Posts: 541
Joined: 2005-05-19 12:06pm

Post by petesampras »

SWPIGWANG wrote:While it would be possible to build a computer that can do this, there is simply no estimate of how hard it would be, and one can just argue that it is not achieved in the universe. In anycase, one could argue against the existence of humans at all if given that level of technology.
It's taken 60 years for computers to play world champion level chess, achieve basic competance in visual and audio recognition, solve undergrad level maths problems, solve logic problems, achieve basic motor skills, etc...

Just 60 years! How long did that take evolution? Extrapolate for another 400-500 years.
User avatar
Vendetta
Emperor's Hand
Posts: 10895
Joined: 2002-07-07 04:57pm
Location: Sheffield, UK

Post by Vendetta »

petesampras wrote:It's taken 60 years for computers to play world champion level chess,
Chess is a system with a limited number of valid legal positions. When it comes to games on a wider scale, computers are still hopeless. They can't, for example, be programmed to play Go with any competence, because it has far more legal positions. (There are more possible legally playable games on a 19x19 Go board than there are atoms in the universe)

Trying to program a computer to run a complicated tactical analysis in the way SF likes to have them do is a hiding unto nothing, because we simply aren't good enough at writing programs for them.

That's why apparently complex RTS behaviours in computers are usually a few simple scripted routines, and in extremely complex games like 4x games they have to start with huge advantages to make any sort of progress at higher levels. And still, anyone who is familiar enough with the algorithms running the computer can fake it out and kill it in ways it's programmers could not have anticipated. (and that would be the first step in military intelligence if computers were running the show. Work out what their possible response range is and act outside it)
petesampras
Jedi Knight
Posts: 541
Joined: 2005-05-19 12:06pm

Post by petesampras »

Vendetta wrote:
petesampras wrote:It's taken 60 years for computers to play world champion level chess,
Chess is a system with a limited number of valid legal positions. When it comes to games on a wider scale, computers are still hopeless. They can't, for example, be programmed to play Go with any competence, because it has far more legal positions. (There are more possible legally playable games on a 19x19 Go board than there are atoms in the universe)

Trying to program a computer to run a complicated tactical analysis in the way SF likes to have them do is a hiding unto nothing, because we simply aren't good enough at writing programs for them.

That's why apparently complex RTS behaviours in computers are usually a few simple scripted routines, and in extremely complex games like 4x games they have to start with huge advantages to make any sort of progress at higher levels. And still, anyone who is familiar enough with the algorithms running the computer can fake it out and kill it in ways it's programmers could not have anticipated. (and that would be the first step in military intelligence if computers were running the show. Work out what their possible response range is and act outside it)
I think you are missing the point. I am not claiming that computers as they are now could replace human 'space' fighter pilots. I am pointing out the tremendous advances computers have made in a very short length of time. 'Space warfare' is likely hundreds of years in our future (if it ever happens), computers, on their current rate of advancement, will be vastly more powerful that those around today.

Also, whilst chess is more limited that go, it still has a combinatorial explosion of game positions which make brute force searches impossible. Chess algorithsm are more cunning than that.

There is still a long way to go, but the rate of advance is staggering. You mentioned game A.I. That too has improved enourmously, even in complex RTS and 4X games. Of course a human is still superior at the moment, but compare the A.I. of warcraft 3 to the original c&c or galactiv civ to the original civ.

Human intelligence remains fairly constant - maybe it is increasing or dcreasing very slowly. Machine intelligence is going up like a rocket.
User avatar
Nyrath
Padawan Learner
Posts: 341
Joined: 2006-01-23 04:04pm
Location: the praeternatural tower
Contact:

Post by Nyrath »

SWPIGWANG wrote:1. Culture reasons: There has been beliefs of board members that war should always be fought by people or else we'd forget the meaning of it. While modern culture attempts to perserve life, a future culture with biological immortality may not have the same view or one may suffer impossible population explosion. The possible culture reasons for this sort of thing can be easily derived from the desire of having manned fighters to begin with. For example, serving could be a rite of passage ritual for a culture.
I refer you to the post immediately before yours.
http://bbs.stardestroyer.net/viewtopic. ... 15#2346618
User avatar
Gunhead
Jedi Council Member
Posts: 1715
Joined: 2004-11-15 08:08am

Post by Gunhead »

It's a bit pointless to assume that computers will continue to improve at the current rate over say 100 years or so. Like the saying goes:"Predicting anything, specially the future, is really hard". Even if we limit our reach to our own solar system, reaching that limit might take say 300 years. If the computers reach their peak in 100 years, people would be travelling between planets using computer tech that's 200 years old.

Now notice the amount of ifs there. No one can prove that computers won't improve at the current rate over the next 100 years or so. But no one can prove that they will.

-Gunhead
"In the absence of orders, go find something and kill it."
-Generalfeldmarschall Erwin Rommel

"And if you don't wanna feel like a putz
Collect the clues and connect the dots
You'll see the pattern that is bursting your bubble, and it's Bad" -The Hives
Enforcer Talen
Warlock
Posts: 10285
Joined: 2002-07-05 02:28am
Location: Boston
Contact:

Post by Enforcer Talen »

If you like the idea of rapid computer advancement, run a search for Ray Kurzwiel. http://www.kurzweiltech.com/aboutray.html
Image
This day is Fantastic!
Myers Briggs: ENTJ
Political Compass: -3/-6
DOOMer WoW
"I really hate it when the guy you were pegging as Mr. Worst Case starts saying, "Oh, I was wrong, it's going to be much worse." " - Adrian Laguna
petesampras
Jedi Knight
Posts: 541
Joined: 2005-05-19 12:06pm

Post by petesampras »

Gunhead wrote:It's a bit pointless to assume that computers will continue to improve at the current rate over say 100 years or so. Like the saying goes:"Predicting anything, specially the future, is really hard". Even if we limit our reach to our own solar system, reaching that limit might take say 300 years. If the computers reach their peak in 100 years, people would be travelling between planets using computer tech that's 200 years old.

Now notice the amount of ifs there. No one can prove that computers won't improve at the current rate over the next 100 years or so. But no one can prove that they will.

-Gunhead
Sure you can't prove for certain one way or the other, that much is obvious. That doesn't mean we shouldn't try and extrapolate what we can from the available evidence.

Computer tech has been improving at an exponential rate for the past 60 years.

Computers already have vastly superior serial speed and precision compared with the human brain.

The human brain is still vastly more complex that any computer.

We thus know that intelligence of the complexity of the brain is possible. It seems reasonable that given, a) it is possible to have a machine with that complexity and b) the exponential increases in the advancements of computer tech, it is likely we will be able to build intelligent machines of the same complexity as the human brain some day.

It is thus reasonable to talk about the possibility of an A.I. with the complexity of the human brain and the speed and precision of modern computers. Such a machine would likely be vastly more competant at any task than a human.

Now, the question of machines with complexity far greater than the human brain or with speed and precision far greater than modern computers, is even more uncertain.

However, if we can match the complexity of the brain with our machines, we will surpuss our brain by virtue of the areas where machines are already superior.
User avatar
GrandMasterTerwynn
Emperor's Hand
Posts: 6787
Joined: 2002-07-29 06:14pm
Location: Somewhere on Earth.

Post by GrandMasterTerwynn »

Gunhead wrote:It's a bit pointless to assume that computers will continue to improve at the current rate over say 100 years or so. Like the saying goes:"Predicting anything, specially the future, is really hard". Even if we limit our reach to our own solar system, reaching that limit might take say 300 years. If the computers reach their peak in 100 years, people would be travelling between planets using computer tech that's 200 years old.

Now notice the amount of ifs there. No one can prove that computers won't improve at the current rate over the next 100 years or so. But no one can prove that they will.

-Gunhead
Even if all advancement in the field of computer science were to completely and irrevocably grind to a halt tomorrow, manned space-fighters would still be a stupid idea. Even now, a computer is likely better at burn calculations than you are. Even now the idiot scripted game AIs we currently have could likely kick your ass piloting a spacecraft modeled on realistic physics.

And in a future with computer-guided missiles or attack drones, they don't really need to be better than a human. They just have to be better than the computers controlling your defenses, since power and room for kinetic slugs will be at a premium aboard your spaceship. This won't be like WW2, where human gunners spray bullets almost willy-nilly at passing enemy fighters. Your railgun turrets will be computer-controlled, since the computer will be better at predicting where an enemy drone will be than you are, and the computer won't panic and empty the turret's magazines in a desperate fit of spray-and-pray.

Sure, a human might be responsible for determining the initial launch pattern and deployment of drone formations. A human is also likely be the one figuring out the best way to draw opposing forces into position to launch the attack. But once the drones and missiles lock on, it'll be time for the humans to get the hell out of the way and be ready for the merde to hit le ventilatuer.
User avatar
Gunhead
Jedi Council Member
Posts: 1715
Joined: 2004-11-15 08:08am

Post by Gunhead »

GrandMasterTerwynn wrote:snipzorzz
All that is true. I was just pointing out there's no need to include super advanced AI into a scifi story/setting. All stories need a certain amount of "It's so because I say so". I'm beginning to wonder how much is the creator of the OP is willing to bend the rules of reality to have his fighters in space.

I loathe the idea of fighters in space. It's completely unrealistic, but more importantly it has the cool factor of a dead snail.

-Gunhead
"In the absence of orders, go find something and kill it."
-Generalfeldmarschall Erwin Rommel

"And if you don't wanna feel like a putz
Collect the clues and connect the dots
You'll see the pattern that is bursting your bubble, and it's Bad" -The Hives
User avatar
frogcurry
Padawan Learner
Posts: 442
Joined: 2005-03-13 06:34am

Post by frogcurry »

The gist of what has been said so far seems to be that there is no way to create a situation where manned spacefighers are needed that doesn't seem to attract some serious ire or is in denial of common sense thinking.

Hence, alternate suggestion, would it be practical to have a story in which some sort of human-equivalent AI is in control of the missle-bus/fighter?

I can think of two flaws so far; 1) You'd probably need self-aware characters but I wouldn't expect a smart military to make such an AI self-aware unless it
was necessary/unintentional, but that seems a bit forced (again), and

2) it would be something that probably dictate a bit of the story and the themes involved (unless you were a very good writer and could avoid the obvious storyline cliches).
User avatar
GrandMasterTerwynn
Emperor's Hand
Posts: 6787
Joined: 2002-07-29 06:14pm
Location: Somewhere on Earth.

Post by GrandMasterTerwynn »

frogcurry wrote:The gist of what has been said so far seems to be that there is no way to create a situation where manned spacefighers are needed that doesn't seem to attract some serious ire or is in denial of common sense thinking.

Hence, alternate suggestion, would it be practical to have a story in which some sort of human-equivalent AI is in control of the missle-bus/fighter?

I can think of two flaws so far; 1) You'd probably need self-aware characters but I wouldn't expect a smart military to make such an AI self-aware unless it
was necessary/unintentional, but that seems a bit forced (again), and

2) it would be something that probably dictate a bit of the story and the themes involved (unless you were a very good writer and could avoid the obvious storyline cliches).
Yes, not to mention the reader isn't apt to care about a story revolving around a computer being all emo because it has to go out and kill other missile drones for a living. As an element of comic relief, sure. As a pillar of the story . . . not as likely to come off very well. And there's a good risk of the story ending up walking straight into a number of cliches, those being A) OMG evil robots r takin over teh world! / r in ur base killin ur d00ds / r goin apeshit and then killin ur d00ds B) OMG evil robots have taken over the world, but one learns to be 'human' by being captured by humans. C) Robot who aspires to be human and becomes human by force of will, and numerous comedic situations arising from robot's ignorance of human nature and customs.

One might suggest an interesting story could be had by taking some person who got cryonically preserved in the 20th and 21st century . . . only to be digitized into a computer intelligence and then being told to go fight enemy drones or missile carriers, or face deletion. Or humans of the time volunteering to become post-human computer intelligences piloting missile drones against the enemy. Though this sort of story would have a number suspension of disbelief problems, not least of which being the question if a civilization could perfectly digitize a person into a computer, then why can't they create an intelligence from scratch and not have to worry about it having human foibles.
User avatar
frogcurry
Padawan Learner
Posts: 442
Joined: 2005-03-13 06:34am

Post by frogcurry »

Oh well. It was worth trying. Guess the future of space warface is big ships and no fighters.
User avatar
SWPIGWANG
Jedi Council Member
Posts: 1693
Joined: 2002-09-24 05:00pm
Location: Commence Primary Ignorance

Post by SWPIGWANG »

Nyrath wrote:The trouble with that kind of solution is it has to be universal. Or the one evil nation that isn't afraid of sentient computers/has no ban on AI research/has technology that progresses will rule the galaxy.
Just because a culture have advance AI doesn't mean the said advanced AI would actually want to be one a ship!

A meat brain is just another computer. In the end, it doesn't matter what kind of brain you put into a ship, as long as it works well.

Now if computers can indeed do all that a human can, we are looking at the overshadowing of humans completely, and ship AI cores are capable of being characters themselves.
GrandMasterTerwynn wrote:Even if all advancement in the field of computer science were to completely and irrevocably grind to a halt tomorrow, manned space-fighters would still be a stupid idea. Even now, a computer is likely better at burn calculations than you are. Even now the idiot scripted game AIs we currently have could likely kick your ass piloting a spacecraft modeled on realistic physics.

Sure, a human might be responsible for determining the initial launch pattern and deployment of drone formations. A human is also likely be the one figuring out the best way to draw opposing forces into position to launch the attack. But once the drones and missiles lock on, it'll be time for the humans to get the hell out of the way and be ready for the merde to hit le ventilatuer.
You are thinking this from the wrong direction in the context of the question. You are thinking that humans should get out of the way of the battle and let the computers do the job because it is nearly as efficient but involve no loss of life.

However, in the context of the question, it is the human that WANTS to be on the ship. The question is whether by placing a human on a ship, it compromises the mission much.

The answer is that it depends, but not always so much as the cost, energy and mass requirements of keeping a crew alive is probably a tiny fraction of running a large fusion reactor and massive plasma torch drive system.

The human is no replacement for computers, it is just an independent intelligence that can solve another set of problems where the AI may not be suited to.
They can't, for example, be programmed to play Go with any competence, because it has far more legal positions.
If it can't be done, how do our brains do it?
Go players are smart. Programmers are not so smart! Okay, joking aside, a game of Go involves alot of trained intutition with massive pattern recognition. Modern computers have no effective pattern recognition algorithm that comes close to a human and is hopeless at Go, where brute force methods are ineffective.

--------------
Consider the modern era: The ICBM is the ultimate weapon of destruction and it requires no human thought. However the military is still dominate by humans. The reason is simple: wars in this social context is fought with limited level of violence and rules of engagement with objectives far more complicated than destruction of enemy forces.

If force of arms is an extension of politics, it makes sense that humans would have to be in the loop. One thing that should be taken note of is that computers can not take responsibility nor would it be allowed non-deterministic action which may be required in fast changing situations, unless it somehow obtains human level of valuation.

While I can understand computers taking over many fields, I think there is one field that this would be difficult, and that is the analysis of humans. It is not due to fundamental computational issues, but the problematic issue of people analyzing itself which may very be "stuck" in social and moral constraints. As an result, one solution would be to place another human and use thousands of years of evolution to approximate an answer out.
Destructionator XIII wrote:I fail to see the point of either: missile bus nor human-equivalent.
Scaling issues of drive systems. If the optimal drive system is far too large to be used just to accelerate one missile, the missile bus would be used.

It is like how ICBM and SLBM have MIRV warheads as opposed to having a dozen smaller missiles. It is more efficient to use one missile as opposed to a dozen.
And IMO that is a deal breaker. I tend to view human and human level AI as being exactly the same, so this would be mass suicide missions. That is a pretty big ethical problem with disposable intelligences, biological or otherwise.
An AI made to die would find it natural. Hell, it is not a big enough ethical problem to stop real people doing it in real life, from the kamikaze of the Japanese to the sucide bombers of the modern Islamic Fundamentalists. The application of sucide intelligence is certainly realistic enough. The moral issue could very well add a dimension to the story.
Post Reply