Fighters in space

[tharkûn]

Human-piloted fighters would be torn to shreds by drones/drone fighters, see the above.
Why? The human has access to smart pilotless missiles. Combat range is at least 5 minutes ToF. The only time maneuver has any meaning is when ToF is low. Hence why the battleships didn't give a damn about maneuvor (crossing the T being the only exception). So your drone can turn faster ... whoop de diddlesquat, at sufficiently long ranges (where REAL WORLD missile combat occurs) this means nothing. The weapon is always more maneuverable than the launch platform.

So long as the engagement takes place at sufficient distance (and the trend here is for greater ranges of engagement) ... there is minimal difference.

Is a database any less stable the more data it has stored in it? Does it develop bugs when you have 10 scenarios stored or when you enter 1,000 scenarios?
Considering that each scenario has to be individually entered? Hell yes. It's HUMAN ERROR whilst coding that is the problem.

Also, little bugs, should they appear, would not be a serious problem.
HAHAHAHAHAHAHA

The zero was a suberb fighter, early in WWII it decimated the American aircraft. Then the Americans noticed a little "bug" in the zero ... during steep dives (or other high speed maneuvors) it wasn't able to target or turn tightly. With the adoption of dive tactics the zero was quickly beat down. A single flaw, properly exploited, is all it takes for real military to down a craft.

The number of superior military equipment, tactics, etc. that have gone down the crapper due to one MINOR flaw is huge.

If I was talking about, say, something like an Operating System or a program designed to run on an operating system, where there are tons of variables and things that could go wrong, I might agree with you. However, we're talking about something that would not have to worry about the dangers of feature creeping like most programs do. Turn on a normal calculator. Do any equation you can do with the input allowed to you. Try to make it crash, try to make it give you a wrong answer. Are there any bugs in it?
Integer overload. I recommend (999^999)^999/(999^999)^999, calculator cannot deal with this. It is piss easy to overload a calculator and make it give you the wrong answer, hell ((1-1E-99)/2-.5 )^99 dies without overloading the calculator. Getting wrong answers from a calculator is child's play. Now given certain input of course the calculator gets it correct, however a battle AI doesn't have this luxury, it has to be prepared for whatever the enemy will throw at it.

Besides I love how you pick a SMALL amount of code and compare it a large amount of code. There were numerous bugs in the original missile guidance systems, and those were nothing but autopilots.

Hah! What garbage. A drone would not be any more or less completely predictable than a human pilot in any significant way.
Really, point me to a computer whose actions are not predicted by its source code. You know it's its optimization, you know it's weighting of solutions, you likely know its seed values. All an AI can do is find the minimum based off its internal algorithms, if you know the alogirthms you can trap it in a local minima while your system is going for global.

Human pilots are less predictable because they don't follow an optimization routine. For instance when Deep Blue beat Kasparov, Deep Blue was simply evaluating board positions, Kasparov was using tactics, strategy, etc. Give Kasparov exact knowledge of how Deep Blue weights said board positions and some time on a supercomputer and I have no doubt he can build a trap.

They can only do so many things in a fighter. Figure out what each of those options are, then set up a simulation based on every possible thing a human pilot could do in a fighter. A lot of those things will be stupid, like ejecting 2.305 seconds into the battle without a vaccuum suit (hey, who would expect someone to do that in a battle?), many will be just plain bad (never shooting at anything the whole battle), some will be decent, a few will be good, even less will be excellent, and the chances of doing something so incredibly spectacular so as to win the battle at the last minute are so rare, unlikely, suicidal, or improbable so as not to make a massive difference. So if the enemy can do simulations with the AI so far as to beat them every time, they can do just the same thing with humans, only there's more to the subject than your example allows.

You can lead an AI down a path. If I can run the AI 1 million times I know what it's first 100 moves in a chess game will be. I can then lead it towards a nice trap based on its minimization path ... offer an attractive local minima and know a better less probable one for your missiles to perform.

So you know every possible response the drones could have to a situation. How does that help you in a battle? If we're talking about a turn-based competition like Chess or Go, then knowing that might help, but in a battle?
Missile battles ARE turn based. Only morons think that engagement ranges are going to decrease. Engagements in missile duels are long range. The enemy fires, you know this from optics (which is hellishly faster than anything else), you take counter measures, if you have AMM's then those impact at about 2/3rds of the engamement distance. Close range combat is DEAD. It has been for years and it will stay that way barring unforseen development.

Space combat will be akin to a missile duel between Aegis cruisers. Going a few knots faster doesn't mean didly squat, turning radius is not important when your weapons are omnidirectional.

The best way to screw an AI is not to predict its tactics (though that is useful), it's to predict how to cheat its friend or foe identification or to trick it into either walking away from its position or not budging from its position. With humans you don't know how they weight the choices for stay here/go afeild after target. With AI you do. You then present a target that the AI will go after and then slip past its position.

With humans you simply don't know. They rarely have detailed plans for deciding these things, and even if they do they often don't follow them.

Why? Human-piloted fighters are slower to change velocity, and thus more predictable. Also, the way humans are trained, they might commonly only learn, a few ways to evade incoming fire, and in the heat of the moment, remember only one or two. Much more predictable than an AI with potentially thousands of randomly selected evasion algorithms. In any event, it's still much harder to shoot down a drone than a human-piloted fighter, not to mention much harder to defend against, as the drone will always be far more accurate with the weapons given it than a human pilot could ever be.
Yes the AI has 1000's of options, all of which are worthless. Evading fire in a missile duel DOESN'T HAPPEN, you take point defense shots, use ECM, and hope for the best. Juking away from a missile is laughable.

The drones superior maneuvering only matter if the engagement is ludicriously short range.

Problem is, you'd have to be able to bribe not just one programmer, but dozens or more.
Not if you are smart. You have one programmer swap an = for a != in a few key locations not likely to come up often, then in the feild you see if it survived and then exploit.

First, obviously, would be a method of classifying the project
I'm sorry I thought you were smart enough to know that is a given. Second off that has ZILCH for bearing on whether or not you get people bribed or coerced (remember putting a gun to somebody, or their family works just as well).

Secondly, the code itself would be double-checked to see how each function works.
Yes so was all of windows 95. So you double check all of the code, errors STILL make it through.

Third, if a captain wanted to kill the Prime Minister with the drones on his ship, he'd be perfectly capable of doing so. They are weapons . They do what they are told. Just as a gun will fire a bullet when you pull the trigger, so will a drone destroy a target if you give the order.
This is a BAD thing. There is a reason why modern WMD's require TWO men (at minimum to launch). More often it takes a good 20-30 people to get weapons off ... NOT one.

The problem would come in if you tried to subvert a number of drones into ignoring the normal chain of command to obey orders from some third party. This can be squashed easily through cryptography.
No, then I just jam the hell out of your communications so your drones can't be ordered to fire.

Adding in a back door through that would be very easy to find and eliminate unless the whole programming team was on the take, in which case you've got a serious problem anyway .
Yes it is the coup scenario you have blatenly ignored. In order for military coups to take place now you need sizable support in terms of manpower. with drones you need a much smaller number.

What if I bribed one of the members of the design team on a human-piloted fighter so that he made some of the systems purposefully faulty? Like if you hit it with a certain amount of a specific type of radiation, all the electronics shut down and the fuel tanks are dumped?

The argument cuts both ways, and the only way you can make it support your position is by arbitrarily making the drone designers idiots and traitors, while not applying the same limitations to your own position.
Not really. My position is having fighters on carriers. If one subsystem is sabotaged then I do in the feild repairs. Yours are stand alone drone missiles which have to be recalled to be repaired. Military equipment regularly has something go hideously wrong, so some bright tech fixes it with duct tape and shouts up the chain of command to get it fixed.

No, they are not. Proper security measures (based on things we can easily do today, mind you) and a competant understanding of programming would prevent the absurd "vulnerabilities" you have claimed drones would have.
Right exponentially increasing the amount of code doesn't increase the number of coding errors. Further reducing the number of people you need to subvert in a coup is a good thing. And reducing it so one captain can order a kill strike against the prime minister is infinately better than having to have 20 people on the same boat.

Nonsense. Regardless, even assuming you could predict every single possible action a drone could make, how would that help in a heated battle if you are using human piloted fighters which are slower moving, slower to react, less likely to hit the target, and prone to making stupid mistakes?
Because battles are largely irrelevant. Compotent militaries try never to fight battles they cannot win easily. You seem to have this mental picture of fighters engaging in dogfights where they can quickly close the distance, where a few xtra g's of transient acceleration will save your ass from a missile ... that ain't gonna happen. Space battles are going to be affairs of lobbing missiles at great distances, letting your missiles try to get through enemy counter measures and using your own counter measures to stop his missiles. The loser is the first guy not to stop a missile, or the first one to run out of missiles.

This is NOT going to be WWII dogfights.

What matters is will the drone target you or not. Will the drone pursue you or not. Convince the drone to follow you away from it's position ... and allow other ships to come through. Convince the ship to stay put, whilst other forces slip around it.

Say a computer chose at random which cup to put the poison into, and let's say, for the sake of argument, that it would leave at least one cup free of poison.
The point is neural nets don't make random decisions. Drones seek minimas based on the weighting criteria you program into them. Truly random decisions are worthless in battle.

Oh, gee, only other programmers can stop them. And I suppose they wouldn't...why? As above, just bribe an engineer to make a subtle yet devestating design flaw in the fighter, it would produce the same effect. Only other engineers can stop him, so you run the same exact risk both ways. This does nothing to support your position.

Sigh because you have technicians in the feild who have been known to fix problems in highly unorthodox manners. Your drone is, however, hard coded. The only way to fix it is to recall it. You also have pilots who say, gee the targetting computer is consistently 2 degrees to the right, I'll aim 2 degrees to the left. Your drone will simply be programmed to shoot 2 degrees to the left.

If you decide to surrender after the point where the trigger is pulled and before the bullet hits you, it's too late .

Which is why it's generally considered a GOOD IDEA to have a system that reaches the point of no return as late in the mission as possible, like say having a human at the firing controls.

How about drone carriers? Regardless, anything that would be able to destroy drones easily would be able to destroy human-piloted fighters even better .
Missile carriers suck. The idea is to have MANY, CHEAP launch platforms, this means fighters or missile boats (i.e. missile frigate). Two missile carriers in a duel means you either win big or you lose big. Two swarms of missile carrying fighters is most likely a draw (normally considered a good thing). All along I have said use fighters to carry smart missiles. This is the most likely outcome because it is where EVERY COMPOTENT MILITARY IS HEADED. Dog fighting fighters are DEAD, it is all about carrying BtH AtA missiles and firing without seeing the other guy.

Hence AI controlled Drone Fighters. Concession accepted.
What are the good points for having drone controlled fighters?
1: faster reaction time
2: better maneuver, higher top acceleration
3: lower mass requirements


to get this you sacrifice:

1: late human control.
2: easy coding on your missiles.
3: decision making capabilities.

Until you have a human level AI, or your human programmers have thought up every eventuality ... humans are better decision makers.

And removing human pilots from the fighters makes them even better. For the record, I don't see any reference made here to using drones as the sole unit of the military. That would be stupid. The point was that they would be better than human-piloted fighters, and that point still stands.

No the initial post read, ".......why fighter when you can missile?

and is there any good reason for space fighters to exist without having physics that make newton roll in his grave?"

Now the initial question is why have fighters? The answer is because they are CHEAP LAUNCH PLATFORMS. They have great operational range (thanks to staging from a carrier), and their is divisibility of force (not possible with cap ships that launch missiles.

So you come down to:
Yes there should be fighters, just control them with AI's who have better decision making capabilities than humans.

I come down to:
Have fighters, staff them with humans unless you are DAMN confident in your programmers and the AI's they build.

A neural net source code bears no resembelance to the compiled and implemented systems. Each drone is differently compiled. All the source code would allow you to do is simulate a "newborn" drone. not a real help to anyone you'd have to do that with each drone. Avoiding whatever countermeasures the drones have.
And how much time does it take to train each net? Considering the number of variables is going to be in the thousands, the number of possible starting datasets in the 10^90? Hell for a neural network you need a global minimum ... that alone will be hell to find. Neural networks cannot even learn the function y = 1/x on the open interval from 0 to 1. It is not just some panacea to say neural networks. You need training data, and loads of it. You need processing time, and loads of that. And finally you need a systems of weights that drops to a global minima ... otherwise you can drop into a local minima.

Congratulations this is an argument for missiles in what way? Dont make me de an ass and sat neutrinos
Because you can make X-ray lasers from missiles, possibly gamma ones. The problem with laser weapons is that it simply becomes a question of who has the most guns with the biggest power supply.

by the time you have finished looking at the display my hundred drones for the cost of your training and fighter and pay and pension etc. have annihalated you. The drones will make better descions than you in around a thousanth the time you do.
Sense when are neural nets cheap? How much time do you plan to dump into training the damn things for huge multivariable functions? How much for error checking and code it for all the functions it inherently can't know. How much do you honestly think training and pensioning a pilot costs? How much do you think the fighter costs? Currently the price tag for a pilot is about 1/100th the cost of the plane (for instance in the F-22), given that you need a MUCH better computer in the drone with far superior decision making capabilities ... forget about it.

The only reason drone craft are cheaper is because they suck in terms of quality. They have far shorter projected lifespan and take numerous shortcuts (i.e. they get to weld lots more stuff than real planes). Most of this cost advantage will go out the window in space. Mass concerns for the human will be minimal.

Show me ANY evidence that you can make an equivalent AI drone (NOT a UAV) at a 100 to 1 price ratio. Current Predators are only 1/8th the cost of the F-22, with NOWHERE near the capabilities.

and can think faster and carries more weapons to mass unit and has a better range of Counter measures.

think faster = worthless it has to think more to get to the same result. The limiting factor is NOT reaction time, it is ToF on your damn missiles.

More weapons/mass ... to a degree, if you don't spend your excess mass suping the engine, providing a bigass computer, etc.

Better countermeasures ... name them.

I've produced a hundred odd drones for that each one using a neural net system. You can defeat one drone. ONE I have a hundred per fighter you have.
BS. At best you have 8 to 1. And that's making the laughable assumption an F-22, at its top rated cost, is equivalent to a Pred UAV. The majority of cost in fighters is in building the damn fighter ... not in the pilot.

good for you. How much good is that when each drone's software is individual to it? but wait you have the source code! congratulations on your drone which knows nothing at all.
How long do you spend on training each drone? How many iterations do you plan to take to train for each every bloody variable? How much more expensive will the computer hardware be?

Besides which the whole idea behind a neural net is to find an approximation to a value ... not find every local minima ... use something like simulated annealing for the latter.

Individualized software is DAMN expensive and time consuming.

]Strategic Division of labour. The extreme security risk in your programmers is worse. I've bribed them to add a piece of code to prevent your fighters deccelerating after a cetain date. You just got X metal coffins. Or to open the hatch of your fighters after a certain date. I can make many more drones per unit than you can fighters and pilots
More BS. There is no real world data supporting hideously cheaper drones, unless they are hideously weaker and stupid. As far as bribing the programmers ... this is why you have MANUEL OVERRIDES. This is where you pull the plug on the computer and do things by hand. This is NOT GOING TO BE STAR TREK, if the computer is buggy you do things MANUALLY.

You can either go the cruise missile route or the AI fighter route. The former is cheap, but is SEVERLY limited. The latter is just as expensive as a fighter.

And a drone can't launch missiles
Will you guys pick a platform and stick to it? Are these drones AI fighters or are they missiles meant to replace fighters?

The initial question is will fighters exist? Fighters are the smallest vessel capable of carrying and launching missiles. Drones place absolute trust in the programmers and the guys giving the drones orders. Drones are also limited by their programming. Currently there are no computer networks that can come close to human level decision capabilities, I doubt they will be here before space fighters.

[Hotfoot]

Human-piloted fighters would be torn to shreds by drones/drone fighters, see the above.
Why? The human has access to smart pilotless missiles. Combat range is at least 5 minutes ToF. The only time maneuver has any meaning is when ToF is low. Hence why the battleships didn't give a damn about maneuvor (crossing the T being the only exception). So your drone can turn faster ... whoop de diddlesquat, at sufficiently long ranges (where REAL WORLD missile combat occurs) this means nothing. The weapon is always more maneuverable than the launch platform.

*sighs* Not simply turn faster, change velocities faster. It has higher overall acceleration rates than a human-piloted craft, would be able to make better use of countermeasures, and so on. You could also afford to fit a drone with offensive beams and cannons, as it would be much more likely to score a damaging hit on a target than a human piloted fighter. Also take in mind that in space velocity is relative. A drone could do a lighting strike against the barely-aware human pilots.

So long as the engagement takes place at sufficient distance (and the trend here is for greater ranges of engagement) ... there is minimal difference.

At longer ranges, the difference in acceleration becomes even more valuable against missiles. Think about it, the missile accelerates towards you, you can accelerate towards the missile. Continue until your relative velocity is too high for the incoming missile to accurately adjust its flight path should you suddenly change velocity. The longer the range of engagement, the more time to react and deploy countermeasures. We've already got lasers that can shoot down artillery shells in midair (albiet under controlled circumstances, for now), it wouldn't be that hard to see a "point defense" laser shooting down an incoming missile in space at distances of even a light-second.

Is a database any less stable the more data it has stored in it? Does it develop bugs when you have 10 scenarios stored or when you enter 1,000 scenarios?
Considering that each scenario has to be individually entered? Hell yes. It's HUMAN ERROR whilst coding that is the problem.

Right. So, for some reason, error checking will have been crippled for no apparent reason. Have you ever even compiled and tested a program before?

Also, little bugs, should they appear, would not be a serious problem.
HAHAHAHAHAHAHA

The zero was a suberb fighter, early in WWII it decimated the American aircraft. Then the Americans noticed a little "bug" in the zero ... during steep dives (or other high speed maneuvors) it wasn't able to target or turn tightly. With the adoption of dive tactics the zero was quickly beat down. A single flaw, properly exploited, is all it takes for real military to down a craft.

You're talking about a mechanical design flaw, not a bug. You're also assume that any little bugs that show up could be exploited to the point where a distinct advantage would be garnered.

The number of superior military equipment, tactics, etc. that have gone down the crapper due to one MINOR flaw is huge.

I guess that depends on your definition of "minor". Got a list, perchance, or some examples of superior military equipment or tactics that were scrapped in such a fashion?

If I was talking about, say, something like an Operating System or a program designed to run on an operating system, where there are tons of variables and things that could go wrong, I might agree with you. However, we're talking about something that would not have to worry about the dangers of feature creeping like most programs do. Turn on a normal calculator. Do any equation you can do with the input allowed to you. Try to make it crash, try to make it give you a wrong answer. Are there any bugs in it?
Integer overload. I recommend (999^999)^999/(999^999)^999, calculator cannot deal with this. It is piss easy to overload a calculator and make it give you the wrong answer, hell ((1-1E-99)/2-.5 )^99 dies without overloading the calculator. Getting wrong answers from a calculator is child's play. Now given certain input of course the calculator gets it correct, however a battle AI doesn't have this luxury, it has to be prepared for whatever the enemy will throw at it.

A normal calculator is only designed to handle a certain amount of numbers. Going beyond that number isn't a bug, it's a design limitation. That's like trying to get a 9mm pistol to fire rounds designed for a .50 caliber Desert Eagle. You grossly misunderstand the problem here.

Besides I love how you pick a SMALL amount of code and compare it a large amount of code. There were numerous bugs in the original missile guidance systems, and those were nothing but autopilots.

Well pardon me for trying to correct your ignorance of coding by starting with the very basics.

Hah! What garbage. A drone would not be any more or less completely predictable than a human pilot in any significant way.
Really, point me to a computer whose actions are not predicted by its source code. You know it's its optimization, you know it's weighting of solutions, you likely know its seed values. All an AI can do is find the minimum based off its internal algorithms, if you know the alogirthms you can trap it in a local minima while your system is going for global.

At best, you might be able to predict its decisions within a fraction of a millionth of a percent. Hardly enough to gain any sort of noticable value from it, and in order even get that, you'd need drones of your own with better computing power than the drones your enemy has.

Human pilots are less predictable because they don't follow an optimization routine. For instance when Deep Blue beat Kasparov, Deep Blue was simply evaluating board positions, Kasparov was using tactics, strategy, etc. Give Kasparov exact knowledge of how Deep Blue weights said board positions and some time on a supercomputer and I have no doubt he can build a trap.

Human pilots are less predicatable? Really? Even if you have the same sort of insight into their training as the AI's programming? Bullshit. A pilot trained at Flight Academy X under Teacher Y will commonly use Z tactics in a fight. We commonly make plans based around how the enemy trains their troops and the methods they are known to commonly use.

You give humans entirely too much credit. Even if Kasparov could lay a trap for Deep Blue given that knowledge, there is no guarantee that it would work, and even if it did, Kasparov would have to have a considerable amount of time beforehand to work out a plan, and then even more time during the game to make it work. Chess is turn-based, so there is no pressure of time, unless you are using an overall time limit, in which case, Deep Blue will still always be able to react faster than Kasparoc. In the realtime arena of a battlefield, reaction time is everything. The human pilots can't spend three minutes thinking how to counter the Drones next move even if they could figure out what it would be in the first place.

They can only do so many things in a fighter. Figure out what each of those options are, then set up a simulation based on every possible thing a human pilot could do in a fighter. A lot of those things will be stupid, like ejecting 2.305 seconds into the battle without a vaccuum suit (hey, who would expect someone to do that in a battle?), many will be just plain bad (never shooting at anything the whole battle), some will be decent, a few will be good, even less will be excellent, and the chances of doing something so incredibly spectacular so as to win the battle at the last minute are so rare, unlikely, suicidal, or improbable so as not to make a massive difference. So if the enemy can do simulations with the AI so far as to beat them every time, they can do just the same thing with humans, only there's more to the subject than your example allows.

You can lead an AI down a path. If I can run the AI 1 million times I know what it's first 100 moves in a chess game will be. I can then lead it towards a nice trap based on its minimization path ... offer an attractive local minima and know a better less probable one for your missiles to perform.

Oh, and somehow the AI would not be able to react to whatever it is that you're doing, based on its programming? Allow me to open your mind to the reality of the situation. You might be able to run the simulation a million times, but the moment you interact with it, you have changed the pattern. An AI does not simply ignore the changes you would enact upon it's arena, be it a chess board or a battlefield in space. And with a random mechanism thrown in, there is no way you can be sure it will take the path you want it to. In the end, its processing speed and technical superiority would still give it the advantage.

So you know every possible response the drones could have to a situation. How does that help you in a battle? If we're talking about a turn-based competition like Chess or Go, then knowing that might help, but in a battle?
Missile battles ARE turn based. Only morons think that engagement ranges are going to decrease. Engagements in missile duels are long range. The enemy fires, you know this from optics (which is hellishly faster than anything else), you take counter measures, if you have AMM's then those impact at about 2/3rds of the engamement distance. Close range combat is DEAD. It has been for years and it will stay that way barring unforseen development.

Oh FFS. Look, turn-based games are an abstraction. Battles happen in real time. There might be a significant lapse in time between firing the missile and the missile hitting the target, but things are still happening during that period of time. I don't see what any of this has to do to help your argument anyway. You're not allowed an infinite amount of time to think about something a computer can calculate in a fraction of a second. You're going off on a huge tangent here, which has nothing to do with drones or human-piloted fighters, and something which I never brought up in the first place.

Space combat will be akin to a missile duel between Aegis cruisers. Going a few knots faster doesn't mean didly squat, turning radius is not important when your weapons are omnidirectional.

Looks like momma didn't teach you about space, boy. Ships don't have speeds, they have acceleration rates. Turning on a dime in space is easy. Why, Earth does it every day! It's changing velocity that makes all the difference in the world, and even a battleship can dodge a missile if the velocities are high enough and the range is long enough.

The best way to screw an AI is not to predict its tactics (though that is useful), it's to predict how to cheat its friend or foe identification or to trick it into either walking away from its position or not budging from its position. With humans you don't know how they weight the choices for stay here/go afeild after target. With AI you do. You then present a target that the AI will go after and then slip past its position.

Predicting a drone's tactics is highly improbably and gains you very little in return. As for cheating the IFF or giving it false orders, I can quite confidently say that I can make it absolutely impossible for you to do that through encryption. And since drone operators would be able to give the drones commands (such as "hold position"), being able to "weight" their choices (which I don't see how you'd do anyway, but to play along) is basically worthless. You seem to be under the impression that the AI would never be telling their command ship/station their status, or able and willing to accept new orders from a verified source, and yet human pilots do that all the time. Why, pray tell, would an AI not be able to do that?

With humans you simply don't know. They rarely have detailed plans for deciding these things, and even if they do they often don't follow them.

So humans are better because they rarely make orders, much less follow them? Right. Call me when you get a clue.

Why? Human-piloted fighters are slower to change velocity, and thus more predictable. Also, the way humans are trained, they might commonly only learn, a few ways to evade incoming fire, and in the heat of the moment, remember only one or two. Much more predictable than an AI with potentially thousands of randomly selected evasion algorithms. In any event, it's still much harder to shoot down a drone than a human-piloted fighter, not to mention much harder to defend against, as the drone will always be far more accurate with the weapons given it than a human pilot could ever be.
Yes the AI has 1000's of options, all of which are worthless. Evading fire in a missile duel DOESN'T HAPPEN, you take point defense shots, use ECM, and hope for the best. Juking away from a missile is laughable.

In atmosphere, where relative velocities are limited, as are manuevering capabilities, yes. In space, however, it's another story entirely. Drones are still harder to hit than human-piloted fighters regardless. Concession accepted.

The drones superior maneuvering only matter if the engagement is ludicriously short range.

Note that I said it is quicker to change velocity. That is primarily due to the fact that it has significantly higher acceleration rates, which would make a difference in a long-range fight. Remember, a missile has to track its target. Something that is attempting to evade is in the position to make the missile react to it. Besides which, the manueverability also means better control and accuracy over any point-defense weapons mounted on the drone.

Problem is, you'd have to be able to bribe not just one programmer, but dozens or more.
Not if you are smart. You have one programmer swap an = for a != in a few key locations not likely to come up often, then in the feild you see if it survived and then exploit.

And chances are it won't have survived, and you'll be dead. If we assume that your covert ops project works, we can assume that mine works, and all your fighters blow up and kill the pilots. You're just as helpless as I am, and you've got dozens if not hundreds of casualities on your side, while I can just simply recover the drones, fix the problem, and use them again later. You've got to train and outfit replacements for all your pilots, refit all of your fighters to fix the design flaw, and pay for the widows/families of your dead pilots. Victory: drones.

First, obviously, would be a method of classifying the project
I'm sorry I thought you were smart enough to know that is a given. Second off that has ZILCH for bearing on whether or not you get people bribed or coerced (remember putting a gun to somebody, or their family works just as well).

Again, cuts both ways, and it's easier to recover from with drones anyway.

Secondly, the code itself would be double-checked to see how each function works.
Yes so was all of windows 95. So you double check all of the code, errors STILL make it through.

Oh, yes, and how many navy computers use Windows 95?

But wait, you want to compare a commercial product released to the public to a potential military product? Fine. Your fighters are all SUVs with Firestone Tires.

Third, if a captain wanted to kill the Prime Minister with the drones on his ship, he'd be perfectly capable of doing so. They are weapons . They do what they are told. Just as a gun will fire a bullet when you pull the trigger, so will a drone destroy a target if you give the order.
This is a BAD thing. There is a reason why modern WMD's require TWO men (at minimum to launch). More often it takes a good 20-30 people to get weapons off ... NOT one.

Who said the drones would be weapons of mass destruction? THEY REPLACE FIGHTERS. Are fighters considered to be weapons of mass destruction? Besides which, as I have repeated time and again previously, there would be drone operators giving the commands, just as there are gunnery officers giving commands to the gun crew or flight control officers give commands to pilots in flight. THE ONLY HUMAN ELEMENT REMOVED IS THE PILOT.

The problem would come in if you tried to subvert a number of drones into ignoring the normal chain of command to obey orders from some third party. This can be squashed easily through cryptography.
No, then I just jam the hell out of your communications so your drones can't be ordered to fire.

1. How would you jam a frickin' laser?
2. Even if you could jam communications, the drones are given orders prior to launch. You couldn't jam that if you wanted to, and the drones will still be able to fire according to their mission parameters and orders.
3. Why wouldn't a ship carrying drones have ECCM to combat your ECM?

Adding in a back door through that would be very easy to find and eliminate unless the whole programming team was on the take, in which case you've got a serious problem anyway .
Yes it is the coup scenario you have blatenly ignored. In order for military coups to take place now you need sizable support in terms of manpower. with drones you need a much smaller number.

Not at all, and you have yet to provide any solid reasons why this would be so. With proper safeguards on the drones, the best you could do is maybe get them to shut down without alerting dozens of red flags during development, and even that would be difficult. Also, I did not ignore your scenario, I have shown how it is irrelevant, and also how it hurts you just as much with manned fighters, if not more so.

What if I bribed one of the members of the design team on a human-piloted fighter so that he made some of the systems purposefully faulty? Like if you hit it with a certain amount of a specific type of radiation, all the electronics shut down and the fuel tanks are dumped?

The argument cuts both ways, and the only way you can make it support your position is by arbitrarily making the drone designers idiots and traitors, while not applying the same limitations to your own position.
Not really. My position is having fighters on carriers. If one subsystem is sabotaged then I do in the feild repairs. Yours are stand alone drone missiles which have to be recalled to be repaired. Military equipment regularly has something go hideously wrong, so some bright tech fixes it with duct tape and shouts up the chain of command to get it fixed.

One subsystem? I'm talking about a subtle design flaw that would cripple or destroy your fighters at any time of my choosing. You couldn't repair it without sending the fighter back to the factory to have it fixed, and that's assuming you even knew it existed. Again, you are arbitrarily assigning limitations to my position while ignoring said limitations for your own position. Brilliant.

No, they are not. Proper security measures (based on things we can easily do today, mind you) and a competant understanding of programming would prevent the absurd "vulnerabilities" you have claimed drones would have.
Right exponentially increasing the amount of code doesn't increase the number of coding errors. Further reducing the number of people you need to subvert in a coup is a good thing. And reducing it so one captain can order a kill strike against the prime minister is infinately better than having to have 20 people on the same boat.[/code]

1. You assume that drones would be used with faulty AIs
2. You assume that drones would be easy to subvert in the design stage
3. You assume that a rogue captain with a starship at his command is somehow more dangerous with drones than with fighters/missiles/really big cannons.
4. You assume that because drones have replaced the pilots of fighters, that they have also replaced the launch crews, maintenance teams, and commanding officers in charge of giving the drones orders.

The Captain of a ship does not control the drones by hand any more than he would fire the cannons of the ship by hand. He tells his officers what to do and then they tell the people below them what to do. It just so happens that the pilots have been replaced by an AI drone. The deck guns on an Iowa do not care what they are shooting at. The shells they fire do not care what they hit. The Captain gives the order, and if the crew complies, the guns fire. If the crew does not comply, the guns do not fire unless the captain somehow manages to get down to the guns, load them, aim them, and fire them himself. Similarly with bombers and modern day fighters, the captain does not directly tell the pilots to bomb their targets, he tells his officers to give the order.

Meanwhile:

1. You assume that human pilots would be able to predict one of potentially millions of possible actions of an AI, much less compensate for them.
2. You assume that fighters would come out of design without any major flaws to their structural or computer systems, and that any problems could be easily detected and fixed before they became serious.

Yeah, that's realistic.

Nonsense. Regardless, even assuming you could predict every single possible action a drone could make, how would that help in a heated battle if you are using human piloted fighters which are slower moving, slower to react, less likely to hit the target, and prone to making stupid mistakes?
Because battles are largely irrelevant. Compotent militaries try never to fight battles they cannot win easily. You seem to have this mental picture of fighters engaging in dogfights where they can quickly close the distance, where a few xtra g's of transient acceleration will save your ass from a missile ... that ain't gonna happen. Space battles are going to be affairs of lobbing missiles at great distances, letting your missiles try to get through enemy counter measures and using your own counter measures to stop his missiles. The loser is the first guy not to stop a missile, or the first one to run out of missiles.

I'll assume you meant "competent" rather than "component". I guess that means that your military would just give up if the other side is using drones instead of fighters then, wouldn't they? Let's review, shall we?

Drones can:

• get into range faster than human fighters can
• get out of range faster than human fighters can (missile dance, good against slower ships)
• knock down incoming missiles better than human fighters can
• evade missiles better than human fighters can
• Intercept ship-killers and human fighters faster and more efficiently than human fighters can
• lay down more accurate and combined firepower better than human fighters can
• use less resources than human fighters
• have longer deployment ranges and times than human fighters
• can deal with close-range combat in the event it should happen better than human fighters can.

In all, drones are far superior to human piloted fighters.

This is NOT going to be WWII dogfights.

No, really? I couldn't tell. Of course, if you've got offensive lasers, then even 300,000 kilometers is short range, but I guess you didn't consider that.

What matters is will the drone target you or not. Will the drone pursue you or not. Convince the drone to follow you away from it's position ... and allow other ships to come through. Convince the ship to stay put, whilst other forces slip around it.

Uh-huh. And you propose to do this how? Just knowing the programming of the drone does not give you magical insight into how to control it.

Say a computer chose at random which cup to put the poison into, and let's say, for the sake of argument, that it would leave at least one cup free of poison.
The point is neural nets don't make random decisions. Drones seek minimas based on the weighting criteria you program into them. Truly random decisions are worthless in battle.

It's not truly random. Assume we use a seed, and you know how the computer will pick the random number. Assume that there's even a 0.00005% chance that it will pick one option more often than the others. How many battles will that win you? What use is it to you? Adding in a random factor is not worthless, according to you, but now it is? Make up your mind, or at least stop applying a double standard to your arguments.

Oh, gee, only other programmers can stop them. And I suppose they wouldn't...why? As above, just bribe an engineer to make a subtle yet devestating design flaw in the fighter, it would produce the same effect. Only other engineers can stop him, so you run the same exact risk both ways. This does nothing to support your position.

Sigh because you have technicians in the feild who have been known to fix problems in highly unorthodox manners. Your drone is, however, hard coded. The only way to fix it is to recall it. You also have pilots who say, gee the targetting computer is consistently 2 degrees to the right, I'll aim 2 degrees to the left. Your drone will simply be programmed to shoot 2 degrees to the left.

Even technicians would be incapable of fixing a serious design flaw. Also, if we are talking about a neural net, gee, it would figure out that its shots are going two degrees to the left of where it intended them to go. It could fix itself with that minor of a problem. That would be ridiculously easy for a neural net (or even just a normal computer program) to correct.

If you decide to surrender after the point where the trigger is pulled and before the bullet hits you, it's too late .

Which is why it's generally considered a GOOD IDEA to have a system that reaches the point of no return as late in the mission as possible, like say having a human at the firing controls.

Humans are unreliable in that regard. Some might shoot the enemy anyway out of spite. If the ship/target surrenders, then the drones can be recalled, just like a fighter. How often, by the way, do fighter pilots tune in to enemy transmissions? How would they know that the enemy has surrendered before a drone would know, exactly?

How about drone carriers? Regardless, anything that would be able to destroy drones easily would be able to destroy human-piloted fighters even better .
Missile carriers suck. The idea is to have MANY, CHEAP launch platforms, this means fighters or missile boats (i.e. missile frigate). Two missile carriers in a duel means you either win big or you lose big. Two swarms of missile carrying fighters is most likely a draw (normally considered a good thing). All along I have said use fighters to carry smart missiles. This is the most likely outcome because it is where EVERY COMPOTENT MILITARY IS HEADED. Dog fighting fighters are DEAD, it is all about carrying BtH AtA missiles and firing without seeing the other guy.

First off, if drones are replacing fighters, then that means that a Drone Carrier would replace a Fighter Carrier, not a Missile Carrier. Thus, your assertion that a Drone Carrier would suck is worthless. I refer you to the points of drone superiority mentioned above.

Hence AI controlled Drone Fighters. Concession accepted.
What are the good points for having drone controlled fighters?
1: faster reaction time
2: better maneuver, higher top acceleration
3: lower mass requirements


to get this you sacrifice:

1: late human control.
2: easy coding on your missiles.
3: decision making capabilities.

The good points are listed above. Drone fighters could carry the same missiles as human-piloted fighters, and in fact could concievably carry more and use them much more effectively. You lose no decision-making capabilities, and you lose no significant late human control. Human-piloted fighters and bombers are still reliant on orders from their carrier before attacking. If communications are lost, they rely on their previous set of orders until contact is re-estabilished. If anything but that happens, it's usually human error on behalf of the pilots, and there is hell to pay afterwards.

Until you have a human level AI, or your human programmers have thought up every eventuality ... humans are better decision makers.

Hence why you have drone controllers to give the drones orders. And you needn't think of every possible situation, just every possible response. Oddly enough, that second list isn't as long as the first, despite your misconceptions.

And removing human pilots from the fighters makes them even better. For the record, I don't see any reference made here to using drones as the sole unit of the military. That would be stupid. The point was that they would be better than human-piloted fighters, and that point still stands.

No the initial post read, ".......why fighter when you can missile?

and is there any good reason for space fighters to exist without having physics that make newton roll in his grave?"

Now the initial question is why have fighters? The answer is because they are CHEAP LAUNCH PLATFORMS. They have great operational range (thanks to staging from a carrier), and their is divisibility of force (not possible with cap ships that launch missiles.

And it quickly turned into "why bother using human-piloted fighters in combat". The only realistic use I can think of for human-piloted fighters is basically police duty, in-system security. Hard combat in small craft like that would be better suited to drones and drone-controlled fighters.

So you come down to:
Yes there should be fighters, just control them with AI's who have better decision making capabilities than humans.

Which, by the time we would have any sort of deep space military, I'd bet that we have the computers and programmers who could handle that easily, given the current rate of development.

I come down to:
Have fighters, staff them with humans unless you are DAMN confident in your programmers and the AI's they build.

Meanwhile you have a gross misunderstanding of almost all the variables involved, in order to keep drones out of combat for as long as possible.

A neural net source code bears no resembelance to the compiled and implemented systems. Each drone is differently compiled. All the source code would allow you to do is simulate a "newborn" drone. not a real help to anyone you'd have to do that with each drone. Avoiding whatever countermeasures the drones have.
And how much time does it take to train each net? Considering the number of variables is going to be in the thousands, the number of possible starting datasets in the 10^90? Hell for a neural network you need a global minimum ... that alone will be hell to find. Neural networks cannot even learn the function y = 1/x on the open interval from 0 to 1. It is not just some panacea to say neural networks. You need training data, and loads of it. You need processing time, and loads of that. And finally you need a systems of weights that drops to a global minima ... otherwise you can drop into a local minima.

I'd say it's safe to assume that there would be a "master" neural net which has all of the essential basics already figured out which could be imprinted into the new neural nets. From there, very little processing time would be needed in comparison. In any event, I don't think that you'd even need a neural net for each drone, really, but that's just me.

Congratulations this is an argument for missiles in what way? Dont make me de an ass and sat neutrinos
Because you can make X-ray lasers from missiles, possibly gamma ones. The problem with laser weapons is that it simply becomes a question of who has the most guns with the biggest power supply.

How is that a problem, exactly? War isn't supposed to be fair, man.

by the time you have finished looking at the display my hundred drones for the cost of your training and fighter and pay and pension etc. have annihalated you. The drones will make better descions than you in around a thousanth the time you do.
Sense when are neural nets cheap? How much time do you plan to dump into training the damn things for huge multivariable functions? How much for error checking and code it for all the functions it inherently can't know. How much do you honestly think training and pensioning a pilot costs? How much do you think the fighter costs? Currently the price tag for a pilot is about 1/100th the cost of the plane (for instance in the F-22), given that you need a MUCH better computer in the drone with far superior decision making capabilities ... forget about it.

How much does a human life cost? Planning on making a clone army anytime soon?

The only reason drone craft are cheaper is because they suck in terms of quality. They have far shorter projected lifespan and take numerous shortcuts (i.e. they get to weld lots more stuff than real planes). Most of this cost advantage will go out the window in space. Mass concerns for the human will be minimal.

Wow, yet another arbitrary limitation for drones. You just don't quit, do you?

Show me ANY evidence that you can make an equivalent AI drone (NOT a UAV) at a 100 to 1 price ratio. Current Predators are only 1/8th the cost of the F-22, with NOWHERE near the capabilities.

There are other factors to consider here. For example:
How long does it take to make a Predator?
How long does it take to train a pilot?

and can think faster and carries more weapons to mass unit and has a better range of Counter measures.

think faster = worthless it has to think more to get to the same result. The limiting factor is NOT reaction time, it is ToF on your damn missiles.

Think more to get to the same result? Funny, it removes a ton of variables that can cloud human judgement and somehow that's a disadvantage?

More weapons/mass ... to a degree, if you don't spend your excess mass suping the engine, providing a bigass computer, etc.

A modern day jet fighter can already handle more Gees than the pilot can. You wouldn't really need to soup up the engines or the frame that much in the first place.

Better countermeasures ... name them.

Point-defense solid-state chemical lasers with computer targeting
Complete knowledge of the guidance systems of all known missiles, and their weaknesses (hey, what goes around comes around, right?), not to mention the ability to store more countermeasures than a human-piloted fighter craft and better judgement as to when to use them for maximum effect (and the reaction time to do so).

I've produced a hundred odd drones for that each one using a neural net system. You can defeat one drone. ONE I have a hundred per fighter you have.
BS. At best you have 8 to 1. And that's making the laughable assumption an F-22, at its top rated cost, is equivalent to a Pred UAV. The majority of cost in fighters is in building the damn fighter ... not in the pilot.

You do not need human pilots for the drones. You are limited by your military enlistment. Drones are limited by industrial capacity. Also, if drones are less likely to be destroyed, that means that during the course of a campaign, you will be outnumbered consistantly more and more as things continue. Maybe not 100 to 1, but enough so that it makes a difference. Also, remember that a drone carrier does not have to allow for pilots or pilot comforts in designing the ship, so a Drone Carrier the same size as one of your Fighter Carriers will be capable of holding many more fighters and drones than you.

good for you. How much good is that when each drone's software is individual to it? but wait you have the source code! congratulations on your drone which knows nothing at all.
How long do you spend on training each drone? How many iterations do you plan to take to train for each every bloody variable? How much more expensive will the computer hardware be?

Well, if you have planetary computers capable of easily decoding all the possible decisions a drone could make...

Probably not very long at all, and it would probably be pretty cheap, too.

Besides which the whole idea behind a neural net is to find an approximation to a value ... not find every local minima ... use something like simulated annealing for the latter.

Individualized software is DAMN expensive and time consuming.

Not in a world where planetary computers like yours exist.

]Strategic Division of labour. The extreme security risk in your programmers is worse. I've bribed them to add a piece of code to prevent your fighters deccelerating after a cetain date. You just got X metal coffins. Or to open the hatch of your fighters after a certain date. I can make many more drones per unit than you can fighters and pilots
More BS. There is no real world data supporting hideously cheaper drones, unless they are hideously weaker and stupid. As far as bribing the programmers ... this is why you have MANUEL OVERRIDES. This is where you pull the plug on the computer and do things by hand. This is NOT GOING TO BE STAR TREK, if the computer is buggy you do things MANUALLY.

Hence the drone operators. There would obviously be other safeguards in place, but I guess you have assumed that everyone proposing drones has the average intelligence of a Starfleet Officer.

You can either go the cruise missile route or the AI fighter route. The former is cheap, but is SEVERLY limited. The latter is just as expensive as a fighter.

Even if it is more expensive than a fighter, it has distinct advantages that make it much better. Much higher efficiency, greater returns, and ZERO loss of human life should it be destroyed. War isn't about dying for your country, it's about making some other poor bastard die for his.

And a drone can't launch missiles
Will you guys pick a platform and stick to it? Are these drones AI fighters or are they missiles meant to replace fighters?

I guess you missed the emoticon, signifying sarcasm in response to your previous statement. There is no alternation of platforms.

The initial question is will fighters exist? Fighters are the smallest vessel capable of carrying and launching missiles. Drones place absolute trust in the programmers and the guys giving the drones orders. Drones are also limited by their programming. Currently there are no computer networks that can come close to human level decision capabilities, I doubt they will be here before space fighters.

The decisions a pilot in a fighter craft makes are extremely limited. The biggest leap would be making a computer program capable of flying a fighter with some degree of skill. Not surprisingly, this is actually easier to do in space, as there are far fewer variables to have to worry about than in atmospheric flight. Beyond that, the decision-making capabilities of a pilot are marginal at best. Whether it's drones or remotely piloted fighters, the military is slowly moving away from having humans in the cockpits of fighters. Drones are simply better because they don't become entirely worthless if communications are lost, while remotely controlled fighters aren't worth a damn without constant two-way communications.

[SirNitram]

Ignoring someone who can't understand that drones are superior to humans(The phrases cost, reaction time, and G-tolerance should be taught to those who disagree with me), I find no reason to employ fighter style platforms at all.

Allow me to illustrate the main reasons:

1) Cost. One large thing will always cost less than many small things, for the simple fact that the one large thing has one large engine(For example), not lots of smaller engines.
2) Range. A larger array for detecting your foe is ultimately going to get better range than many small ones. Note: Many small ones will improve accuracy within range, so maybe it would help to have sensor drones.
3) Sustainability. A capital ship fully loaded with supplies can always outlast a short range drone.

[Hotfoot]

Ignoring someone who can't understand that drones are superior to humans(The phrases cost, reaction time, and G-tolerance should be taught to those who disagree with me), I find no reason to employ fighter style platforms at all.

Allow me to illustrate the main reasons:

1) Cost. One large thing will always cost less than many small things, for the simple fact that the one large thing has one large engine(For example), not lots of smaller engines.
2) Range. A larger array for detecting your foe is ultimately going to get better range than many small ones. Note: Many small ones will improve accuracy within range, so maybe it would help to have sensor drones.
3) Sustainability. A capital ship fully loaded with supplies can always outlast a short range drone.

Well, as I said in the beginning of the thread, police and security duties would be peachy fine for fighters. For pure combat, drones would be better than fighters, if you absolutely had to have them (probably good for defensive forces, if nothing else).

Drones could have multiple uses in a fleet. Sensor/scout drones, high accuracy mobile point defense drones (to assist the normal ship-based PD), and as smart decoys. Also, drones would be invaluable in collecting any battle salvage, as if something goes wrong, you just lose a drone.

[SWPIGWANG]

Lets just say if you can subvert my programers, I can subvert people that make your fusion reactors.

Anyway, I guess the future of war would involve both sides throwing lots of processing power at all levels to get a advantage. So when combat starts, one would have computers calculating what the other side would do and another calculating what the out side would do if they know what you are going to do based on the 1st decision and this goes on and on to where the point of dimished returns.

But no, a drone is no more predictable then a manned fighter, as they can be reloaded software on the fly and rely on random number generators. The limits are only the storage space of the drone and the communication speed, both are increasing exponentially as of now. Humans on the other hand, could be extreamly predictable as they have nature emotions and knee jerk responses that can be exploited. While no Ai can model emotions as of now, it is not impossible and the greatest military minds in history is relient on knowing what the enemy will do.

I'll just have a few hundred programming teams and have programs swap on the drones at real time if needed. So if my mainframe sees that the current drone software is no good, upload a new one.

And lack of optimization means it is inferior

[tharkûn]

*sighs* Not simply turn faster, change velocities faster. It has higher overall acceleration rates than a human-piloted craft, would be able to make better use of countermeasures, and so on. You could also afford to fit a drone with offensive beams and cannons, as it would be much more likely to score a damaging hit on a target than a human piloted fighter. Also take in mind that in space velocity is relative. A drone could do a lighting strike against the barely-aware human pilots.
Not really. Lightning fast for anything of mass is still piss slow for optics.

At longer ranges, the difference in acceleration becomes even more valuable against missiles. Think about it, the missile accelerates towards you, you can accelerate towards the missile. Continue until your relative velocity is too high for the incoming missile to accurately adjust its flight path should you suddenly change velocity.
You are thinking of modern conventional weapons. Weapons in space , by virtue of the vacuum in space, need to be area effect with no pressure wave from the surrounding air. You are looking at nuclear weapons. No amount of juking is going to save you from one of those suckers going off close range. If you have an AI capable of flying a fighter and hitting targets on the fly ... then you can but the same AI in the missile.

Shooting down missiles in flight is a reason why human don't lag hideously behind drones in space combat. Have an Aegis type system with a limited AI and you are good to go. The pilots need only activate the weapons leaving the targetting to the computers. The fraction of a second the drone gains per volley is nonexistant given the ToF.


Right. So, for some reason, error checking will have been crippled for no apparent reason. Have you ever even compiled and tested a program before?
Nothing more a few kb in length. Error checking happens, but the DARPA still can't get completely bug free software for some applications. Hell the computer system that manages fissile material was recently found to have a potentially catostrophic glitch.


Are you honestly claiming no bugs survive compilation and code checking?


You're talking about a mechanical design flaw, not a bug. You're also assume that any little bugs that show up could be exploited to the point where a distinct advantage would be garnered.
Mechanical bug the craft performs poorly at high speeds.

Software bug the targeting controls perform poorly at high speeds.

It takes one thing your design team doesn't think of for the whole thing to be faulty.

I guess that depends on your definition of "minor". Got a list, perchance, or some examples of superior military equipment or tactics that were scrapped in such a fashion?
Nuclear AtA missiles. Defeated when the Soviets realized they existed and altered SOP's. More powerful versions were nixed due to EMP concerns.

Ground beater anti-mine tanks (tanks with an extension off the front that beat the ground with heavy chains to set off mines in their path) ... somebody thought about using a time delay on the mines.

US rifles automatically ejected the clip from the gun when it was empty. This often hit something hard and gave off a distinct sound which served as signal to the enemy that US soldier was out ammo and was reloading ... needless to say this was changed later.

At Zama the Romans found that while Elephants were nigh to impossible to stop, so they instead used trumpets, yells, etc. to frighten the elephants and generally convince them that they were in danger. The elephants broke and ran either back into the Carthagian battlelines or down pre-arranged corrridors. Thus ended the reign of the elephant in Mediterrainian war.

In each of these cases a single flaw like an empty clip hitting the ground or elephants prone to panicing was ruthlessly exploited by the enemy.

A normal calculator is only designed to handle a certain amount of numbers. Going beyond that number isn't a bug, it's a design limitation.
You said, and I quote:
try to make it give you a wrong answer
I did so.
How?
By exploiting something beyond its design. Does that matter? No. The enemy doesn't give a flying frik what you designed it for, their job is to find what overloads it, what makes it crash, and when is it predictable. The enemy is also quite good at it. If design limitations are an issue then you need to be damn sure your AI fighter doesn't have them, that means designing for every possibility.

At best , you might be able to predict its decisions within a fraction of a millionth of a percent. Hardly enough to gain any sort of noticable value from it, and in order even get that, you'd need drones of your own with better computing power than the drones your enemy has.

Whatever. If your code is basically random what is so superior? Why bother with an AI if you can just throw dice? Yes I know you'd need more computing power, I'd estimate at least two orders of magnitude ... which is why when you get the code for a drone you take to a mainframe and study it there.

Human pilots are less predicatable? Really? Even if you have the same sort of insight into their training as the AI's programming? Bullshit. A pilot trained at Flight Academy X under Teacher Y will commonly use Z tactics in a fight. We commonly make plans based around how the enemy trains their troops and the methods they are known to commonly use.
Yes and these change without having to rip equipment apart and recode. Remember this is hardcoded in so the only way to change the coding is to physically change it. Where you trained is of much less value in a war than what theatre you are in, what duties do you perform (bomber, fight-fighter, fighter-bomber), who you are in wing with, etc. You could get that data for every fighter, but then you have to know which planes they are in. Unless each drone is unique you need only get the data for one drone.

You give humans entirely too much credit. Even if Kasparov could lay a trap for Deep Blue given that knowledge, there is no guarantee that it would work, and even if it did, Kasparov would have to have a considerable amount of time beforehand to work out a plan, and then even more time during the game to make it work. Chess is turn-based, so there is no pressure of time, unless you are using an overall time limit, in which case, Deep Blue will still always be able to react faster than Kasparoc. In the realtime arena of a battlefield, reaction time is everything. The human pilots can't spend three minutes thinking how to counter the Drones next move even if they could figure out what it would be in the first place.
No they just need to set the counter measures on auto (like Aegis of today) and let them take the point shots.

Oh, and somehow the AI would not be able to react to whatever it is that you're doing, based on its programming? Allow me to open your mind to the reality of the situation. You might be able to run the simulation a million times, but the moment you interact with it, you have changed the pattern.
Duh, which is why on the first 500,000 you don't interact. On the next 500,000 you simulate the interaction and look for a solution that converges to your desired goal.

And with a random mechanism thrown in, there is no way you can be sure it will take the path you want it to. In the end, its processing speed and technical superiority would still give it the advantage.
Random mechanisms only work if the degree of randomity is sufficient not to render the AI stupid, but enough that its weighting functionscan find the global minima.

Look, turn-based games are an abstraction. Battles happen in real time.
You are forgetting the ToF. In real time you can only fire so many missile/sec ... this limited by the hardware, not the software. It takes so many fractions of a section to relay this information on light pulses. The enemy then has so many minutes to engage counter measures. These have a discrete rate at which they can occur and again are limited by the hardware. It takes another chunk of time for the countermeasures to hit the incoming missile.

At sufficiently long engagement ranges it is a series of moves. Shaving 10 seconds off of reaction time means squat with multiple minute ToF. Split second relfexes go the way of the dinosaur. As engagement distances increase, reaction times become less and less important.

Looks like momma didn't teach you about space, boy. Ships don't have speeds, they have acceleration rates.
It is called an analogy Aegis cruisers do have speeds.

Turning on a dime in space is easy
It's also useless with omnidirectional weapons.

ven a battleship can dodge a missile if the velocities are high enough and the range is long enough.
Only if you engage outside the optimal range. Dodging missiles ain't gonna happen, no matter what you due to that BB there is nothing that will make it faster than a missile. Nothing that makes it more maneouverable. AMM's and point defense shots will save you, juking won't.

Predicting a drone's tactics is highly improbably and gains you very little in return. As for cheating the IFF or giving it false orders, I can quite confidently say that I can make it absolutely impossible for you to do that through encryption.
With or without quantum computing? Cheating FoF is just a matter of looking at how it determines friend or foe ... what digitized inputs illicit a friend response, and which illicit a foe response. You can blackbox the entire inner workings, but give me a working modle and a bigger computer ... I can find way to change it.

You seem to be under the impression that the AI would never be telling their command ship/station their status, or able and willing to accept new orders from a verified source, and yet human pilots do that all the time. Why, pray tell, would an AI not be able to do that?
Because you have said it will be HARDCODED in. When presented with problem of hacking the response was to hardcode the drone and restrict access to physical only. Or are you guys changing the rules yet again?

If you can reprogram FoF remotely, hello hacking potential.

So humans are better because they rarely make orders, much less follow them? Right. Call me when you get a clue.
No because humans are less predictable, but more flexible than your hardcoded drone. Humans can't be trusted to be tricked. Humans if tricked can change their FoF procedures without having to return from the feild to have their brains yanked out and rewired.

In atmosphere, where relative velocities are limited, as are manuevering capabilities, yes. In space, however, it's another story entirely. Drones are still harder to hit than human-piloted fighters regardless. Concession accepted.
You are going to be using THERMONUCLEAR missiles. Your average shot is going to have a kill radius in the KILOMETRES. Juking is NOT an option. In space you also don't have civillians who dislike it when you set off nukes overhead. Your craft needs to juke say a kilometre in a tenth of second ... it AIN'T gonna happen. The only reason to use conventional weapons on earth is their secondary effects ... which don't occur in space.

Note that I said it is quicker to change velocity. That is primarily due to the fact that it has significantly higher acceleration rates, which would make a difference in a long-range fight. Remember, a missile has to track its target. Something that is attempting to evade is in the position to make the missile react to it.
How many ORDERS OF MAGNITUDE are you talking about? If it is less than one forget about juking the nuke.


Besides which, the manueverability also means better control and accuracy over any point-defense weapons mounted on the drone.

Somebody doesn't get the idea about point defenses. Point defenses are single point to single point shots. The weapons mounts are flexible and preferably omnidirectional. With advanced AMM's the relative orientation of your ship is completely irrelevant.


And chances are it won't have survived, and you'll be dead.
Right a single embedded error with documentation that says the code is correct, but isn't? Get real. Have you ever tried to find an error the programmer was trying to hide? Today we can't even catch errors that occur by mistake, let alone intentional ones.


If we assume that your covert ops project works, we can assume that mine works, and all your fighters blow up and kill the pilots
Oh right cancel my order instead of something simple I'll change my order to magically make the drones head back to your base and open fire

The fact of the matter is the bigger the code is, the harder it is to debug. The more code your debuggers have to go through the more easily a malicious error can be introduced and hidden. How many years do you plan spend debugging?

Something minor can slip through, rigging the fighter to blow is not going to get magically harder on a drone. I went for something simple, but if you want to play worst case scenario drones are far more vunerable if they get programmed to attack home.

Again, cuts both ways, and it's easier to recover from with drones anyway.
Not when you have to replace all that hardcode and debug a massively greater amount of hardware. With a space fighter the option for manual control always becons ... so does ripping out a specific indepentant computer system and replacing it with a different modle. With a wholly integrated, hardcoded drone you have to remove and debug the whole thing.

But wait, you want to compare a commercial product released to the public to a potential military product? Fine. Your fighters are all SUVs with Firestone Tires.
I'd take that trade any day of the weak. A few hundred accidents LINKED to the tires ... given MILLIONS of tires compared to an OS that crashes regularly.

The more complicated the code, the more it is expected to, the more likely it is to be buggy.


Who said the drones would be weapons of mass destruction? THEY REPLACE FIGHTERS. Are fighters considered to be weapons of mass destruction?
In space? Hell yes. Given the ToF you are going to need nuclear level weapons, the enemy has to much time to react to allow for any conventional warhead (which has crap for effect in a vacuum) to get proximity kills. Also unlike earth a fixed target in space (a free floating space station for at a lagrange point for instance) can have unlimited armor so the only weapons capable of blowing that can be used.

A thermonuclear equipped fighter will most certainly be a WMD.

Besides which, as I have repeated time and again previously, there would be drone operators giving the commands, just as there are gunnery officers giving commands to the gun crew or flight control officers give commands to pilots in flight.
Which are susceptible to hijacking and signal blocking. Yes I know you intend to encrypt everything to hell and back, it still can be hijacked if you have the right data or a computer capable of quantum processing.

In a human piloted fighter you can keep fighting, abeit less effectively, if you are cut off from the carrier.

1. How would you jam a frickin' laser?

Place a mirror between the source and drone. Bath the drone in laser pulses that blow the carrier signal into oblivion. I personally want to know how you intend to ensure the drone receives this laser as it is juking away from nuclear weapons. Is the entire hull a laser receiver?

This also provides excellent means for tracking the carrier, find out where the lasers are coming from ... just look for the excited hydrogen in the sea of colder hydrogen.

. Even if you could jam communications, the drones are given orders prior to launch . You couldn't jam that if you wanted to, and the drones will still be able to fire according to their mission parameters and orders.

Which goes directly back to them needing the decision capabilities as good or better than a human. It goes back to the fact that at these distances the advantages of the AI are minimal.

3. Why wouldn't a ship carrying drones have ECCM to combat your ECM
Because it doesn't have the RANGE. Power goes down with r squared. If I'm jamming a drone at its distance 1 from me, and you are 10 ... you need 100 bloody times as much ECCM as I need ECM. If its 1:50 you need 25,000 times as much power.

Not only that but ECCM on a carrier is asking for it to get slaughtered. ECCM acts as a giant homing beacon, something perfect for swarms of missiles to be targeted against.

Normally carriers try to keep a low profile in actual combat, not show everyone exactly where the biggest target is.


Not at all, and you have yet to provide any solid reasons why this would be so. With proper safeguards on the drones, the best you could do is maybe get them to shut down without alerting dozens of red flags during development, and even that would be difficult. Also, I did not ignore your scenario, I have shown how it is irrelevant, and also how it hurts you just as much with manned fighters, if not more so.

Let's say you have design team of 50 programmers with unlimited access. You have 5 people in the top brass. They decide to coup. 55 people then can bury whatever the hell they like in the code and nobody is wiser. One day the brass tells the ships to launch for training/inspection exercises and to use some set of conditions that was preset triggers the drones do ignore everything the coup leaders want it to ignore. Those fighters that are left in the hangers or not armed suffer complete AI failure. 55 people with access to numerous WMD systems, completely decimating the others.

Now what happens with pilots? The brass have to recruit pilot by pilot, these pilots aren't centralized like you programmers so its harder to bribe/threaten them all. Once you get say 50 pilots in on it you still have to deal with with the thousands of other pilots.

It is a bad idea to allow a few people control over all the major weapons.

One subsystem? I'm talking about a subtle design flaw that would cripple or destroy your fighters at any time of my choosing. You couldn't repair it without sending the fighter back to the factory to have it fixed, and that's assuming you even knew it existed. Again, you are arbitrarily assigning limitations to my position while ignoring said limitations for your own position. Brilliant.
No that is the price you pay for having a completely integrated AI ultimately in control of everything and hardcoding the whole thing.

Subtle design flaws are MUCH harder to pull off when the computer is not ultimately in charge. When individual systems have individual computers and the human can sever any connections ... it is much harder to sabotage. Think about it instead of bathering, why isn't the American nuclear arsenal controlled by computer? Why do you think that might not be a good idea here?

1. You assume that drones would be used with faulty AIs
Loads of military precedent there. The military never, never feilds equipment with bugs in it

You assume that drones would be easy to subvert in the design stage
The more code the harder it is to check. Giving the AI ultimate control means it IS easier to subvert. AI's are not modular, once a peice of its code is subverted you need to check EVERYTHING ELSE. How many years does it take to completely debug gigs of code?

You assume that a rogue captain with a starship at his command is somehow more dangerous with drones than with fighters/missiles/really big cannons.
Yes because with pilots the captain needs to order the fighters to scramble, the deck crew needs to arm them, the pilots need to obey unlawful orders. Generally speaking it takes 20 people.

4. You assume that because drones have replaced the pilots of fighters, that they have also replaced the launch crews, maintenance teams, and commanding officers in charge of giving the drones orders.

No I assume that you have one man giving the drone orders ultimately. There is a reason why you want 2 man crews, both have to be nuts.

1. You assume that human pilots would be able to predict one of potentially millions of possible actions of an AI, much less compensate for them.

No because an AI isn't frikking random. They don't need to plan for millions they need to figure out how the AI handles FoF and then cheat that ... or does your AI randomly open fire on friendlies?

You assume that fighters would come out of design without any major flaws to their structural or computer systems, and that any problems could be easily detected and fixed before they became serious.
No I'm assuming they are modular in design and that computer flaws can be worked manually in a pinch.

The bloody fact of the matter is:
Can drones have structural flaws? Yes
Can human fighters have structural flaws? Yes
Can drones have general computer flaws? Yes
Can human fighters have general computer flaws? Yes
Can drones have AI flaws? Yes
Can human fighters have AI flaws? NO

quit dicking around with well you might have structural flaw/sabotage/whatever that is directly equivalent to an AI error. You have ALL THE SAME BLOODY PROBLEMS, plus that fact that an integrated AI ultimately holds decision power.

get into range faster than human fighters can
Irrellevant, optics gives the humans a heads up, with omnidirectional weapons the humans can engage just as easily.

get out of range faster than human fighters can (missile dance, good against slower ships)
Useless against faster missiles. The missile is always faster than the platform.

knock down incoming missiles better than human fighters can

BS alert both sides use omnidirectional computer targeted countermeasures. The seconds gained are nothing in terms of ToF.

Intercept ship-killers and human fighters faster and more efficiently than human fighters can

Is directly defeated by:

have longer deployment ranges and times than human fighters
If you are father out, you have longer intercept times. Exactly how much faster is your drone?

have longer deployment ranges and times than human fighters
Yes and no. You can deploy longer/farther ... but only if it is completely autonomous or isn't jammed.

can deal with close-range combat in the event it should happen better than human fighters can.

This is drones major advantage ... and it is generally worthless.

No, really? I couldn't tell. Of course, if you've got offensive lasers, then even 300,000 kilometers is short range, but I guess you didn't consider that.
And where are going to pack your powersupply for this weapon? Isn't that in the gigawatts?

Uh-huh. And you propose to do this how? Just knowing the programming of the drone does not give you magical insight into how to control it.
Right whatever. So it is playing dice again?

It's not truly random. Assume we use a seed, and you know how the computer will pick the random number. Assume that there's even a 0.00005% chance that it will pick one option more often than the others. How many battles will that win you? What use is it to you? Adding in a random factor is not worthless, according to you, but now it is? Make up your mind, or at least stop applying a double standard to your arguments.

Double standard? Whatever let's look at what you have taken for granted:
1. That your drone can identify FoF ... cannot be done now.
2. That an equally advanced computer cannot dupe the first (remembering the first has to allow for idiot civillians).
3. A drone that can quickly outthink a human even with millions of variables to cull through. Computers can barely beat humans at chess today ... and chess is an EASY problem for a computer ... one that has no open bounds, one where everything is in the rules.
4. A drone that can receive orders, not be jammed or hacked, and is hardcoded.

Humans are not random, they are merely not as predictable. The data needed to predict humans requires far more intelligence and is far less useabl for modling. AI source code HAS to come digitized, it has to accept readily digitizable inputs, and it doesn't have things like childhood experiences clouding it. Computers have randomity or predictability. Humans are not random nor are they predictable, they appear random.

Your random number generator is CYCLICAL. Deriving the random number values generated is not beyond the ability of computation. If you AI is random why bother with the AI, just use dice.

Even technicians would be incapable of fixing a serious design flaw. Also, if we are talking about a neural net, gee, it would figure out that its shots are going two degrees to the left of where it intended them to go. It could fix itself with that minor of a problem. That would be ridiculously easy for a neural net (or even just a normal computer program) to correct.
Really and it would know this training data how? Does it automatically go back to training mode whenever it encounters something not encoded in its banks?

Neural nets take make approximations. They need to know both the input and the desired output. Eventually (assuming the problem is actually tractable) they are accurate enough to call it. However if you start presenting cases outside its ability to predict ... it needs more training data.

Even technicians would be incapable of fixing a serious design flaw. Also, if we are talking about a neural net, gee, it would figure out that its shots are going two degrees to the left of where it intended them to go. It could fix itself with that minor of a problem. That would be ridiculously easy for a neural net (or even just a normal computer program) to correct.
Because the enemy flies a white flag, because the enemy cuts into their frequency, because the enemy spells out we surrender in the sand.

First off, if drones are replacing fighters, then that means that a Drone Carrier would replace a Fighter Carrier, not a Missile Carrier. Thus, your assertion that a Drone Carrier would suck is worthless. I refer you to the points of drone superiority mentioned above.

Sigh drones can refer to UAVs, smart missiles, even drone capships. Drone fighters are the best solution using drones. You still need to trust your AI, you still need to be damn sure you can't be easily hacked and jammed. The thread starts out will their be fighters, the answer is YES. You are argueing should their be unmanned fighters, remote controlled fighters (which magically cannot be hacked or jammed), or human piloted fighters.

You lose no decision-making capabilities, and you lose no significant late human control. Human-piloted fighters and bombers are still reliant on orders from their carrier before attacking. If communications are lost, they rely on their previous set of orders until contact is re-estabilished. If anything but that happens, it's usually human error on behalf of the pilots, and there is hell to pay afterwards.
Yes and they magically can't be hacked. Oh wait say encryption ten times fast and it all goes away magically.

And it quickly turned into "why bother using human-piloted fighters in combat". The only realistic use I can think of for human-piloted fighters is basically police duty, in-system security. Hard combat in small craft like that would be better suited to drones and drone-controlled fighters.
Hard combat is better done by smart missiles. If your fighter is really nothing more than a launching point for missiles ... which it is becoming ... then the range gets to the point where your drone has no significant advantage.

Which, by the time we would have any sort of deep space military, I'd bet that we have the computers and programmers who could handle that easily, given the current rate of development.
Deep space military will be here as soon as space colonies exist.

Meanwhile you have a gross misunderstanding of almost all the variables involved, in order to keep drones out of combat for as long as possible.
Right and which one of us thinks that a laser ranged at 100,000 km is viable?

Face the facts neural nets are not the be all end all of programming there are numerous simple things they can't deal with. You have ZIP for a garuntee they will ever be combat worthy. They introduce the possibility for hacking and jamming to screw them over more royally than human control.

Their major advantage is they are faster than a human fighter, but slower than a missile.

With omnidirectional weapons, long time of flights, and thermonuclear weapons I'm not seeing huge savings here, what an aditional few g's of thrust and maybe 200 kg in mass savings? All while picking up increased room for error in the code, increased hacking/jamming potential, and the ability for a small number of people to take over all the fighters.

I'd say it's safe to assume that there would be a "master" neural net which has all of the essential basics already figured out which could be imprinted into the new neural nets. From there, very little processing time would be needed in comparison. In any event, I don't think that you'd even need a neural net for each drone, really, but that's just me.

I think a neural net for each drone is stupid, just do one and stamp ... it works.

How is that a problem, exactly? War isn't supposed to be fair , man.
[/b]
Because with long range laser weapons you need HUGE power supplies. In which case your ships will be as slow as whales in any event.

How much does a human life cost? Planning on making a clone army anytime soon?
War is a bloody mess, it will remain so. Saving pontential hundreds of lives whilst risking more with hijackable drones? I think not.

Wow, yet another arbitrary limitation for drones. You just don't quit, do you?
No that is real world price comparison. Did you not notice the direct comparison between the predator and the F-22 ... that IS the only reason real world drones are cheaper they are piss slow, with worse weapons, and fly at low altitudes.

There are other factors to consider here. For example:
How long does it take to make a Predator?
How long does it take to train a pilot?
How long does it take to train the operator, he he is to perform as well as a pilot he needs as much training (more because the stupid ones will survive their mistakes more frequently).

Think more to get to the same result? Funny, it removes a ton of variables that can cloud human judgement and somehow that's a disadvantage?
It adds a ton more it has to discount in overhead processing. Computers can barely beat top chess players and chess is OPTIMIZED for computers ... its a damn simple calculation. It will require exponentially more computation to expand your dataset to battlefeild conditions.

A modern day jet fighter can already handle more Gees than the pilot can. You wouldn't really need to soup up the engines or the frame that much in the first place.
And a the tolerance of a modern fighter pilot for g's keeps going up. The highest g load is still up around 80 transient.

Point-defense solid-state chemical lasers with computer targeting
Complete knowledge of the guidance systems of all known missiles, and their weaknesses (hey, what goes around comes around, right?), not to mention the ability to store more countermeasures than a human-piloted fighter craft and better judgement as to when to use them for maximum effect (and the reaction time to do so).
Yes modern human piloted craft don't ever use computer targetting.

There is didly squat a drone can do that a human with proper targeting computers can't do.

You do not need human pilots for the drones. You are limited by your military enlistment. Drones are limited by industrial capacity. Also, if drones are less likely to be destroyed, that means that during the course of a campaign, you will be outnumbered consistantly more and more as things continue. Maybe not 100 to 1, but enough so that it makes a difference. Also, remember that a drone carrier does not have to allow for pilots or pilot comforts in designing the ship, so a Drone Carrier the same size as one of your Fighter Carriers will be capable of holding many more fighters and drones than you.

Look folks it is the bait and switch. First you had the drone carrier that had humans flying these things remotely, drone control officers. Now you have a carrier with no pilots (a pitfully small number of crew on any carrier ... maybe 100 out of thousands) that doesn't need space for the pilots ... wahoo you can what 3 heads and 25 private quarters from the ship? That adds room for 5% more fighter and mass for a fraction of it? Those drone operators add nothing to ship's compliment, the fact that the mass of crew quarters and support are not likely to MASS anywhere near a single fighter is completely irrelevant.

Well, if you have planetary computers capable of easily decoding all the possible decisions a drone could make...

Probably not very long at all, and it would probably be pretty cheap, too.

Neural nets work because they RETURN THE SAME SOLUTION. If you have a neural net learn the function f(x,y,z,a...) and you have another one learn the function f(x,y,z,a...) and you ask each one to find a minima with tolerance of +/- delta they should ALL RETURN THE SAME ANSWERS. In order to make them react in a significantly different manner you need to have completely different functions governing their behavior ... which means you need to train each one individually with different data sets.

Further you don't need to know everything the drone will do in every circumstances, only those that it will do in a given scenario you want to know about.

Hence the drone operators. There would obviously be other safeguards in place, but I guess you have assumed that everyone proposing drones has the average intelligence of a Starfleet Officer.
Drone operators lack PHYSICAL ACCESS to the systems. Pilots don't. You put a small box next to the pilot with a sequence of circuit breakers. If the computer goes to hell the pilot pulls the fuse for the system effected. If a drone operator senses his AI's are going loopy he can only send EM pulses and pray the drone receives them and isn't too far gone to act on it.

Even if it is more expensive than a fighter, it has distinct advantages that make it much better. Much higher efficiency, greater returns, and ZERO loss of human life should it be destroyed. War isn't about dying for your country, it's about making some other poor bastard die for his.
It also has the added headaches of more code to potentially have errors, more code to be subverted, and the hijacking/jamming problem.

Zero loss of life is a good thing, but zealous pursuit of it is a good way to handicap yourself in war.

The decisions a pilot in a fighter craft makes are extremely limited. The biggest leap would be making a computer program capable of flying a fighter with some degree of skill. Not surprisingly, this is actually easier to do in space, as there are far fewer variables to have to worry about than in atmospheric flight. Beyond that, the decision-making capabilities of a pilot are marginal at best. Whether it's drones or remotely piloted fighters, the military is slowly moving away from having humans in the cockpits of fighters. Drones are simply better because they don't become entirely worthless if communications are lost, while remotely controlled fighters aren't worth a damn without constant two-way communications.
Technically speaking drones are RC fighters. All modern UAV's are drones, and are remote controlled.

I seriously doubt humans will move out of the cockpit anytime soon. The added headaches of cyberwarfare, computer bugs, coupled with ever increasing missile mass, size, speed, and engagement ranges will make the fighter nothing more than a point to launch missiles from.

Really I have yet to see an argument you have put forward that supports human occupied carriers and not just droning out the entire force.

[tharkûn]

Damn that was long let me ask the drone advocates for a few guestimates:
1. How fast will a drone processor be?
2. How much code (mega/giga/tera/whatever bytes) will it be?
3. How much faster will a drone be (i.e. how many m/s/s will a human fly, how many will a drone)? How much more fuel will it carry for these speeds (faster engines being less efficient than slow ones normally)?
4. What size armament will your drone be carrying?
5. What is a typical range of engagement?
6. What data security measures will you be taking (hardcoding, encryption)?
7. How redundant will your drone be?
8. How is a drone superior to a long range AI missile from say a missile frigate?
9. Will the carrier require a human crew?
10. How much mass/space (profile only) will you save going drone?

Please no pulling numbers out of your ass to be cute, make a guess or say you have no clue.

[GrandMasterTerwynn]

.......why fighter when you can missile?

and is there any good reason for space fighters to exist without having physics that make newton roll in his grave?

There are only a few realistic purposes for manned space fighters.

A) Interdiction of near-planet civilian traffic. It consolidates your forces into a space station and it's complement of space fighters. This allows you to free up capital ships for more nefarious uses.

B) Expand your options when deploying missiles. Your carriers would stay outside your enemy's weapons range and deploy their fighters. These fighters could carry a handful of anti-capship missiles. Sure their defenses wouldn't be able to withstand any serious offense, but their size and maneuverability would make them harder to hit. As a result, they could sneak missiles up to your enemy.

C) And with B, this point tends to follow: Interceptors to destroy enemy fighters before they can inflict the nasty sorts of damage that B implies.

Otherwise, building a small vessel with enough armament to threaten at least some capital vessels and having a very small crew . . . a very expensive and mostly preposterous proposition.

[SWPIGWANG]

jeeze, why can't mike come in and delive the imperial smack down, since someone is brute forcing here.
Dodging
1. You can out run missile's fight envelope if you can take the acceleration. (and humans can't take it)

2. You can still dodge lasers and rail guns what not

3. Better strategic mobility due to higher acceleration tolerance

4. You are assuming nukes are going to render passive defense useless, which is not going to be the case as a bit of extra jinking means bigger warhead will be needed to ensure a kill and vice versa. You don't give away any advantage when you can have it.

Useless against faster missiles. The missile is always faster than the platform.

Oh a torpedo is faster than a SSN....oops it is not

And turning takes insane amounts of energy with newtonian physics in space, so one could easily move to a spot where the missile can't reach due to energy constrains.

Command

No because humans are less predictable, but more flexible than your hardcoded drone. Humans can't be trusted to be tricked. Humans if tricked can change their FoF procedures without having to return from the feild to have their brains yanked out and rewired.

Humans betray, computers don't
Humans have other motives, computers don't (causing over stating kills and such)
Humans surrender, and if you are the french space force....
Computers can be reprogramed faster than re-educated

If you can reprogram FoF remotely, hello hacking potential.

Try hacking into quantum communication. Oops you CAN'T BY THE LAWS OF PHYSICS. No, you can't hijack or intercept quantum communication with any physics we know, and having a quantum decoder would not help.

The weapons mounts are flexible and preferably omnidirectional. With advanced AMM's the relative orientation of your ship is completely irrelevant.

Preferable != always
And sometimes you need big guns spinal to defeated armored missiles. (big for a small drone that is)

I went for something simple, but if you want to play worst case scenario drones are far more vunerable if they get programmed to attack home.

I can program your nukes in missiles to blow too. :rolleyes: That is like saying guns are useless because gun powder blows when you burn them.

Place a mirror between the source and drone. Bath the drone in laser pulses that blow the carrier signal into oblivion. I personally want to know how you intend to ensure the drone receives this laser as it is juking away from nuclear weapons. Is the entire hull a laser receiver?

You might as well say you can prevent ICBM lauches by placing a fifty ton steel block on the launch hatch. That notion is absurd because if would be magnitudes easier to blow up the drone than blocking the comm that might be relyed from any other friendly ship in the area from any angle anywhere. Also the laser can be widened to ensure reception. As for interception, I can rely the signal so that your ships are not anywhere near the path the beam passes though, and quantum communication deals with that anyway.

and do you even know how ECCM works?

55 people with access to numerous WMD systems, completely decimating the others.

Oh no, the designers of the ICBM lanuch systems decides to coup and attacks with a newly constructed base while disabling all others!

My ass, that is the problem with human organization not technological one. Besides, even with manned figthers I can just program to shut down the fly by wire system or engines with the same result.

Who says I'll only have one team of programmers? I'll probably have dozens and if any is compromised I'll just switch.

Subtle design flaws are MUCH harder to pull off when the computer is not ultimately in charge. When individual systems have individual computers and the human can sever any connections ... it is much harder to sabotage.

Yes, because a spitfire has a human in it, it can dive and not suffer from fuel cut off in 1940. Anyway, remote control is no less controlled than manned ship, unless you systems fails to the point that the pilot have to physical shoot the computer to pieces.

Think about it instead of bathering, why isn't the American nuclear arsenal controlled by computer?

And it isn't? Only the command structure is manned, the actual weapon itself are all automatic.

The more code the harder it is to check. Giving the AI ultimate control means it IS easier to subvert. AI's are not modular, once a peice of its code is subverted you need to check EVERYTHING ELSE. How many years does it take to completely debug gigs of code?

fdisk C:
format C:
xcopy D:\ C:\

No I assume that you have one man giving the drone orders ultimately. There is a reason why you want 2 man crews, both have to be nuts.

And that is why you want 500 man commanding a SSBN. Opps you don't.

No because an AI isn't frikking random. They don't need to plan for millions they need to figure out how the AI handles FoF and then cheat that ... or does your AI randomly open fire on friendlies?

or does your pilots randomly open fire on friendlies?

There is only a finite number of logical actions in any given sistuation, and given decent programing, a drone can attempt and of those with complete unpredicability. Inlogical actions should be attempted by only idiots like stupid pilots.

No, really? I couldn't tell. Of course, if you've got offensive lasers, then even 300,000 kilometers is short range, but I guess you didn't consider that.
And where are going to pack your powersupply for this weapon? Isn't that in the gigawatts?

Nuke pumped X-ray lasers your ASS.

1. That your drone can identify FoF ... cannot be done now.

I can and your fighter can't without computerized IFF system.

3. A drone that can quickly outthink a human even with millions of variables to cull through. Computers can barely beat humans at chess today ... and chess is an EASY problem for a computer ... one that has no open bounds, one where everything is in the rules.

It beat the best human there was at the time. If beat ALL humans. Unless your pilots are all the top 0.0001% of the human population in terms of skill, you don't even have a chance at all. And chess is far more strategic while less time intensive than tactial combat, so it actually favors humans. Try playing checkers with the computer, or worst, quake 3 on hardest.

Humans are not random, they are merely not as predictable.

By throwing a few qubits or merely use a random number generator using outside info sources (eg. senser input) a drone can be completely unpredictable. Sure you can say that if you knew all the seed values, you could predict it, but you can't because the seed values are completely random and unpredictable, so no amount of modeling will allow you to predict it.

How long does it take to train the operator, he he is to perform as well as a pilot he needs as much training (more because the stupid ones will survive their mistakes more frequently).

Less time because newbies don't die before they learn the basics.

It adds a ton more it has to discount in overhead processing. Computers can barely beat top chess players and chess is OPTIMIZED for computers

MY ASS. Chess is a human game, not a computer one. If you want a game that is optimized for computers it would be a RTS running at 10,000,000 times the speed of what we normally play. (w00t lose to 6 ling in 1 micro second)

In order to make them react in a significantly different manner you need to have completely different functions governing their behavior ... which means you need to train each one individually with different data sets.

And that is faster than growing a human.

Further you don't need to know everything the drone will do in every circumstances, only those that it will do in a given scenario you want to know about.

and that is a advantage when you are a human?

The added headaches of cyberwarfare, computer bugs, coupled with ever increasing missile mass, size, speed, and engagement ranges will make the fighter nothing more than a point to launch missiles from.

why fighters? why not just make long range missiles and call it a military?

[The Yosemite Bear]

I am trying to remember the name of the book, but I am recalling one Sci-Fi series where they got around the problems with fighter;s and physics/lifesupport/feeding. By transfering the pilot's conciousness into the spacecraft. Making the person an Electronic Ghost inside the ship.

It went really well into the French Canadian & Northeast Native American myths about Wandering/Nomad Ghosts too.

[Hotfoot]

*sighs* Not simply turn faster, change velocities faster. It has higher overall acceleration rates than a human-piloted craft, would be able to make better use of countermeasures, and so on. You could also afford to fit a drone with offensive beams and cannons, as it would be much more likely to score a damaging hit on a target than a human piloted fighter. Also take in mind that in space velocity is relative. A drone could do a lighting strike against the barely-aware human pilots.
Not really. Lightning fast for anything of mass is still piss slow for optics.

There's a difference between being able to see the shot that's coming and being able to do anything about it before it kills you.

At longer ranges, the difference in acceleration becomes even more valuable against missiles. Think about it, the missile accelerates towards you, you can accelerate towards the missile. Continue until your relative velocity is too high for the incoming missile to accurately adjust its flight path should you suddenly change velocity.
You are thinking of modern conventional weapons. Weapons in space , by virtue of the vacuum in space, need to be area effect with no pressure wave from the surrounding air. You are looking at nuclear weapons. No amount of juking is going to save you from one of those suckers going off close range. If you have an AI capable of flying a fighter and hitting targets on the fly ... then you can but the same AI in the missile.

And in the vacuum of space, there is no air to produce a shockwave, drastically reducing the area of effect on a nuke. And if you're going to put fighter drone AI on a missile, then that missile has to be large enough to handle the computer system, and it's also non-recoverable, making it much more expensive than the drones fighters. In which case, it still doesn't provide a valid case for human-piloted fighters. In fact, you only further validate my point that manned fighters are worthless in space combat. If one side uses AI Drones, the other side needs to use it to try and break even, otherwise they will get slaughtered.

Shooting down missiles in flight is a reason why human don't lag hideously behind drones in space combat. Have an Aegis type system with a limited AI and you are good to go. The pilots need only activate the weapons leaving the targetting to the computers. The fraction of a second the drone gains per volley is nonexistant given the ToF.

Never mind the numerous other advantages, including the missile dance. Drones would be able to dive in just far enough to launch a missile you couldn't hope to escape from, the fly right back out of the maximum effective range of your missiles. Sure, you could send extra-fuel anti-capship missiles after them, but then you've got fewer to use on their capital ships, don't you?

Right. So, for some reason, error checking will have been crippled for no apparent reason. Have you ever even compiled and tested a program before?
Nothing more a few kb in length. Error checking happens, but the DARPA still can't get completely bug free software for some applications. Hell the computer system that manages fissile material was recently found to have a potentially catostrophic glitch.

I never said the software would be completely bug free, but it would be within acceptable limits. Nothing near the scale you seem to think. Even if every one out of ten drones has a total systems failure and goes dead in space, they're still better than manned fighters by a large margin.

Are you honestly claiming no bugs survive compilation and code checking?

Nope, just no distasterous, terrible, civilization-ending bugs like you seem to think there would be.

You're talking about a mechanical design flaw, not a bug. You're also assume that any little bugs that show up could be exploited to the point where a distinct advantage would be garnered.
Mechanical bug the craft performs poorly at high speeds.

Software bug the targeting controls perform poorly at high speeds.

It takes one thing your design team doesn't think of for the whole thing to be faulty.

Yes, that's right. Double standard all the way, why stop now?

I guess that depends on your definition of "minor". Got a list, perchance, or some examples of superior military equipment or tactics that were scrapped in such a fashion?
Nuclear AtA missiles. Defeated when the Soviets realized they existed and altered SOP's. More powerful versions were nixed due to EMP concerns.

Defeated when the other side found out and covered their operating procedures? How does that make the technology flawed in any way? You're grasping at straws.

Ground beater anti-mine tanks (tanks with an extension off the front that beat the ground with heavy chains to set off mines in their path) ... somebody thought about using a time delay on the mines.

You have a very loose definition of "flaw", don't you?

US rifles automatically ejected the clip from the gun when it was empty. This often hit something hard and gave off a distinct sound which served as signal to the enemy that US soldier was out ammo and was reloading ... needless to say this was changed later.

So the rifle went down the crapper and was never used again, or was the design refined? So far, none of these fit the description you gave...

At Zama the Romans found that while Elephants were nigh to impossible to stop, so they instead used trumpets, yells, etc. to frighten the elephants and generally convince them that they were in danger. The elephants broke and ran either back into the Carthagian battlelines or down pre-arranged corrridors. Thus ended the reign of the elephant in Mediterrainian war.

Okay, that's one example, from the ancient world. I suppose I'll give you that one...

In each of these cases a single flaw like an empty clip hitting the ground or elephants prone to panicing was ruthlessly exploited by the enemy.

Note: You said military equipment and tactics which have gone down the crapper (i.e. scrapped) due to one minor flaw. So far, you have presented one solid example of this from the Ancient World. Everything else does not fit the profile.

A normal calculator is only designed to handle a certain amount of numbers. Going beyond that number isn't a bug, it's a design limitation.
You said, and I quote:
try to make it give you a wrong answer
I did so.
How?
By exploiting something beyond its design. Does that matter? No. The enemy doesn't give a flying frik what you designed it for, their job is to find what overloads it, what makes it crash, and when is it predictable. The enemy is also quite good at it. If design limitations are an issue then you need to be damn sure your AI fighter doesn't have them, that means designing for every possibility.

Okay, let's try this again. In a battle, there are a limited number of potential actions at any given time. This limit is imposed by the abilities of the unit and reality. A tank cannot make bombing runs, an A-10 in flight cannot go hull down, an Infantryman cannot rip a tank to shreds with his bare hands. The possibilities might be quite large, but they are ultimately finite.

You are, quite frankly, gloating because you have proven that a tank cannot drop bombs on enemy targets from 30,000 feet. Does this make tanks a flawed design, because they cannot do things that they were not designed to do? Of course not.

At best , you might be able to predict its decisions within a fraction of a millionth of a percent. Hardly enough to gain any sort of noticable value from it, and in order even get that, you'd need drones of your own with better computing power than the drones your enemy has.

Whatever. If your code is basically random what is so superior? Why bother with an AI if you can just throw dice? Yes I know you'd need more computing power, I'd estimate at least two orders of magnitude ... which is why when you get the code for a drone you take to a mainframe and study it there.

First you claim that you can predict every move a drone would make, then when a little variety is thrown in, a little random element, you claim it makes the drone worthless. Let me spell it out for you further then. If two courses of action would provide roughly equal success, which would you choose? In the case of the AI, if some courses of action provide equal probability of working (this is especially true in evasion), then it would randomly decide which one to take. Thus, you could not predict anything beyond what would be expected by chance. If you manage to force it into a situation where only one option is available, you've expended far more resources in doing so than you would ever have to do with a human-piloted fighter. Additionally, if we are using neural nets, neural nets learn by trying new things, and hence you would never be able to accurately predict the actions to the point where you have a significantly higher advantage over drones than humans.

Human pilots are less predicatable? Really? Even if you have the same sort of insight into their training as the AI's programming? Bullshit. A pilot trained at Flight Academy X under Teacher Y will commonly use Z tactics in a fight. We commonly make plans based around how the enemy trains their troops and the methods they are known to commonly use.
Yes and these change without having to rip equipment apart and recode. Remember this is hardcoded in so the only way to change the coding is to physically change it. Where you trained is of much less value in a war than what theatre you are in, what duties do you perform (bomber, fight-fighter, fighter-bomber), who you are in wing with, etc. You could get that data for every fighter, but then you have to know which planes they are in. Unless each drone is unique you need only get the data for one drone.

Human pilots tend to die, you forget that. When they die, all their expertise and experience dies with them. Most pilots are not stellar aces, despite your longing for them to be, and are usually taught their habits by (shock) where they were trained to fly. And if the military schools are standarized, well, then all the easier. Regardless, being able to predict your enemy's actions has been around for thousands of years. Humans are not some sort of wildly unpredictable force in combat. Quite the opposite, in fact. There might be a very few notable exceptions that prove the rule, but one Red Baron does not a war win.

You give humans entirely too much credit. Even if Kasparov could lay a trap for Deep Blue given that knowledge, there is no guarantee that it would work, and even if it did, Kasparov would have to have a considerable amount of time beforehand to work out a plan, and then even more time during the game to make it work. Chess is turn-based, so there is no pressure of time, unless you are using an overall time limit, in which case, Deep Blue will still always be able to react faster than Kasparoc. In the realtime arena of a battlefield, reaction time is everything. The human pilots can't spend three minutes thinking how to counter the Drones next move even if they could figure out what it would be in the first place.
No they just need to set the counter measures on auto (like Aegis of today) and let them take the point shots.

Well, if AEGIS-type systems work perfectly, then I suppose missiles are completely worthless, aren't they? Ah well, moot point then, does nothing to help your position.

Oh, and somehow the AI would not be able to react to whatever it is that you're doing, based on its programming? Allow me to open your mind to the reality of the situation. You might be able to run the simulation a million times, but the moment you interact with it, you have changed the pattern.
Duh, which is why on the first 500,000 you don't interact. On the next 500,000 you simulate the interaction and look for a solution that converges to your desired goal.

Heh, you must have a lot of free time on your hands then. First of all, you'd need a lot more than just 500,000 simulations with your interaction to get even the faintest idea of what to do. Even then, you're assuming that with the technical superiority of the drones that you would be able to find a way to win in the first place, much less a reliable way to win. And you'd have to figure out how they'd fare against forces X with forces Y at their command at location Z, each time doing anywhere from 1,000,000 to 1,000,000,000 simulations for each to get any sort of meaningful data, and then you've got to do it again for every possible combination of forces and locations that could possibly ever happen. We're talking about approaching infinity type stuff here to gather the sort of statistics that would even be marginally useful.

And with a random mechanism thrown in, there is no way you can be sure it will take the path you want it to. In the end, its processing speed and technical superiority would still give it the advantage.
Random mechanisms only work if the degree of randomity is sufficient not to render the AI stupid, but enough that its weighting functionscan find the global minima.

Yeah, so? You're not exactly blowing any holes in the argument here. Do you think that the excluded middle is the natural state of the universe? Wait, I'm not a Republican, so I must be a Democrat, right?

Look, turn-based games are an abstraction. Battles happen in real time.
You are forgetting the ToF. In real time you can only fire so many missile/sec ... this limited by the hardware, not the software. It takes so many fractions of a section to relay this information on light pulses. The enemy then has so many minutes to engage counter measures. These have a discrete rate at which they can occur and again are limited by the hardware. It takes another chunk of time for the countermeasures to hit the incoming missile.

It. Still. Happens. In. Real. Time. Turn-based games do not have a time limited imposed on them.

At sufficiently long engagement ranges it is a series of moves. Shaving 10 seconds off of reaction time means squat with multiple minute ToF. Split second relfexes go the way of the dinosaur. As engagement distances increase, reaction times become less and less important.

Except when dealing with weapons that travel at 300,000 km/s. Whoops, never mind, lasers don't exist in your world, do they? Because if they do, then all of a sudden split-second timing becomes important again, doesn't it? So does evasion. You know, dodging the barrel?

Looks like momma didn't teach you about space, boy. Ships don't have speeds, they have acceleration rates.
It is called an analogy Aegis cruisers do have speeds.

It's a piss-poor analogy then. Get a better one that suits the discussion at hand.

Turning on a dime in space is easy
It's also useless with omnidirectional weapons.

Your omnidirectional weapons have limited effective range and can be shot down hundreds of thousands of kilometers away.

ven a battleship can dodge a missile if the velocities are high enough and the range is long enough.
Only if you engage outside the optimal range. Dodging missiles ain't gonna happen, no matter what you due to that BB there is nothing that will make it faster than a missile. Nothing that makes it more maneouverable. AMM's and point defense shots will save you, juking won't.

It all depends on relative velocities, like I said. Note the "if".

Predicting a drone's tactics is highly improbably and gains you very little in return. As for cheating the IFF or giving it false orders, I can quite confidently say that I can make it absolutely impossible for you to do that through encryption.
With or without quantum computing? Cheating FoF is just a matter of looking at how it determines friend or foe ... what digitized inputs illicit a friend response, and which illicit a foe response. You can blackbox the entire inner workings, but give me a working modle and a bigger computer ... I can find way to change it.

IFF can be determined through an encrypted signature, every ship and fleet has their own. Without knowing each ship or fleet's private key you wouldn't be able to fake their IFF, and all you need is a public key to decrypt it. For the record, there is a form of encryption which is impossible to break even with quantum computers, and the technology for that is already far ahead of the struggling quantum computer prototypes. And going by the golden rule of cryptography, even if you know the cryptosystem I'm using, you can't do a damn thing about it without the key. Any decent cryptosystem is designed with that in mind. You can't do anything about it. There is crypto which is hard to break, which, even with the biggest computer you have, could take you months if not years, and then there's crypto that is impossible to break. If the first form is not secure enough, the second form can easily be used, and you are left high and dry.

You seem to be under the impression that the AI would never be telling their command ship/station their status, or able and willing to accept new orders from a verified source, and yet human pilots do that all the time. Why, pray tell, would an AI not be able to do that?
Because you have said it will be HARDCODED in. When presented with problem of hacking the response was to hardcode the drone and restrict access to physical only. Or are you guys changing the rules yet again?

If you can reprogram FoF remotely, hello hacking potential.

What the hell are you talking about? You can't attack the operating system of a computer by scanning a picture, what makes you think you could hack a drone's operating system by sending noise down it's microphone? A drone can easily be given orders remotely without altering the operating system in any way. There is also no mention of remotely altering the IFF/FoF system, JUST GIVING THE DRONE ORDERS. In any event, it's a moot point if any form of strong or unbreakable encryption is used, but the fact of the matter is that the drone's operating system has no direct remote connections, and cannot be reprogrammed remotely.

So humans are better because they rarely make orders, much less follow them? Right. Call me when you get a clue.
No because humans are less predictable, but more flexible than your hardcoded drone. Humans can't be trusted to be tricked. Humans if tricked can change their FoF procedures without having to return from the feild to have their brains yanked out and rewired.

Now you're just not making any sense at all. Either reword the point you were trying to get out or give it up.

In atmosphere, where relative velocities are limited, as are manuevering capabilities, yes. In space, however, it's another story entirely. Drones are still harder to hit than human-piloted fighters regardless. Concession accepted.
You are going to be using THERMONUCLEAR missiles. Your average shot is going to have a kill radius in the KILOMETRES. Juking is NOT an option. In space you also don't have civillians who dislike it when you set off nukes overhead. Your craft needs to juke say a kilometre in a tenth of second ... it AIN'T gonna happen. The only reason to use conventional weapons on earth is their secondary effects ... which don't occur in space.

If you are travelling at a high enough relative velocity, why, yes, you CAN evade a missile Virginia. Since you have no puny human to complain when pulling over 100m/s/s, or say 200m/s/s or maybe even more, it becomes much more of a possibility. Meanwhile, humans are going to be limited to 60m/s/s most of the time, or maybe 90m/s/s if they really want to push it. Thus, humans are much easier to hit than drones. You need more manueverable missiles to hit drones than humans, concession accepted.

Note that I said it is quicker to change velocity. That is primarily due to the fact that it has significantly higher acceleration rates, which would make a difference in a long-range fight. Remember, a missile has to track its target. Something that is attempting to evade is in the position to make the missile react to it.
How many ORDERS OF MAGNITUDE are you talking about? If it is less than one forget about juking the nuke.

Tell you what, you give me some hard data on what the radius of your nuke would be, and I'll be able to figure out just what would be needed to evade it, hmm? Let's start with the following:

Weapon yeild
warhead mass
overall missile mass
missile fuel
missile guidance system (is it dumb? Is it smart? Is it a drone AI?)

Besides which, the manueverability also means better control and accuracy over any point-defense weapons mounted on the drone.

Somebody doesn't get the idea about point defenses. Point defenses are single point to single point shots. The weapons mounts are flexible and preferably omnidirectional. With advanced AMM's the relative orientation of your ship is completely irrelevant.

Missiles are omnidirectional weapons in that if the missile tube is not facing the target, the missile will adjust itself and go after it anyway. Logically speaking, a mounted dumbfire weapon cannot be omnidirectional. When it comes to having a highly accurate "Point Defense" frickin' laser beam, it helps to be able to steady your ship for the shot, so that you can hit something that is very far away. Someone doesn't get the idea about point defenses himself, if he doesn't think that a direct-fire weapon is more accurate from a platform that can quickly be made relatively stable to the target.

And chances are it won't have survived, and you'll be dead.
Right a single embedded error with documentation that says the code is correct, but isn't? Get real. Have you ever tried to find an error the programmer was trying to hide? Today we can't even catch errors that occur by mistake, let alone intentional ones.

Here you go again, with the excluded middle. Because some programmers can't catch errors that occur by mistake, NO programmers can ever catch errors the occur by mistake? And that in a program where a primary concern would be a back door program, nobody would ever think to look for one, or know what one would possibly loook like? Sure.

If we assume that your covert ops project works, we can assume that mine works, and all your fighters blow up and kill the pilots
Oh right cancel my order instead of something simple I'll change my order to magically make the drones head back to your base and open fire

Fine, then I space the pilots and have rudimentary AIs blow the fuck out of them then you. Same effect. Anything you can do I can do, so long as you're having your little masturbatory fantasy about perfectly hidden codes and flaws, I can have mine with the same effects, if not worse. Concession accepted, manned fighters have no discernable advantages over drones here.

The fact of the matter is the bigger the code is, the harder it is to debug. The more code your debuggers have to go through the more easily a malicious error can be introduced and hidden. How many years do you plan spend debugging?

The security algorithms, in comparison to the rest of the AI, are ridiculously easy to make secure. You do not need to pour over the entire damn thing to make sure that the security works the way that it was intended to work. You do not pour over your accounting software's source code when you want to find holes in your firewall. Sounds like you're suffering from the same method of programming that brought us the Y2K bug. It would be a very simple measure to add a killswitch to the drones, so that if they did start attacking friendlies or acting against orders, the drone simply goes dead in space, to be recovered later to see just what the problem was, report the causes, and avoid such possible problems until it gets "fixed" in the batch of drones. Old drones are then sent back as new drones are supplied, and then refitted with improved systems so that the drone carriers always have a full payload.

Something minor can slip through, rigging the fighter to blow is not going to get magically harder on a drone. I went for something simple, but if you want to play worst case scenario drones are far more vunerable if they get programmed to attack home.

Hey, guess what, it's not going to get magically easier either. Whatever you can think of as a worst case scenario for drones, I can make even worse for human-piloted fighters, just on the fact that in addition to whatever you have the drones do, all the pilots die. You can't win, move on.

Again, cuts both ways, and it's easier to recover from with drones anyway.
Not when you have to replace all that hardcode and debug a massively greater amount of hardware. With a space fighter the option for manual control always becons ... so does ripping out a specific indepentant computer system and replacing it with a different modle. With a wholly integrated, hardcoded drone you have to remove and debug the whole thing.

See above.

But wait, you want to compare a commercial product released to the public to a potential military product? Fine. Your fighters are all SUVs with Firestone Tires.
I'd take that trade any day of the weak. A few hundred accidents LINKED to the tires ... given MILLIONS of tires compared to an OS that crashes regularly.

How many people did Windows 95 have a direct hand in killing, exactly?

The more complicated the code, the more it is expected to, the more likely it is to be buggy.

Well no shit. Just means it will take longer to develop a reliably stable version. Hey, guess what, more complicated airplanes present more enigineering hurdles than simpler ones! I mean, anyone can design a glider, but from there, do you realize how many kinks we'll have to work out to make a hypersonic aircraft? Well gee, I guess we'll never be able to do it. I mean, if just one person creates a design flaw that makes one plane out of hundred crash, it will be worthless. Boo-hoo, boo-hoo.

Who said the drones would be weapons of mass destruction? THEY REPLACE FIGHTERS. Are fighters considered to be weapons of mass destruction?
In space? Hell yes. Given the ToF you are going to need nuclear level weapons, the enemy has to much time to react to allow for any conventional warhead (which has crap for effect in a vacuum) to get proximity kills. Also unlike earth a fixed target in space (a free floating space station for at a lagrange point for instance) can have unlimited armor so the only weapons capable of blowing that can be used.

Er, when relative velocities are in the multiples of kilometers per second, and short range for effective point defense is in the thousands of kilometers, unless those nukes have fucking huge areas of effect, like, in the hundreds of kilometers, it's not going to be significant difference over conventional warheads.

A thermonuclear equipped fighter will most certainly be a WMD.

So every pilot is packing nukes? Why the hell would you send human-piloted fighters out into that in the first place? They'd either get blown to hell by your nukes-o-doom or not able to do a damn thing against the PD of a capship. Hell, a capship's PD would probably vape a fighter from a few thousand kilometers out. It sure as hell wouldn't be able to evade, and it can't be carrying much protection.

Besides which, as I have repeated time and again previously, there would be drone operators giving the commands, just as there are gunnery officers giving commands to the gun crew or flight control officers give commands to pilots in flight.
Which are susceptible to hijacking and signal blocking. Yes I know you intend to encrypt everything to hell and back, it still can be hijacked if you have the right data or a computer capable of quantum processing.

Signal blocking, yes, in a limited fashion, which is why the drones are capable of making decisions on their own. Hijacking, however, would be impossible with proper security measures. There is encryption that even quantum computing can not break. I'm surprised you even know about quantum decryption without knowing about quantum cryptography.

In a human piloted fighter you can keep fighting, abeit less effectively, if you are cut off from the carrier.

Same with a drone. As I've said numerous times before.

1. How would you jam a frickin' laser?

Place a mirror between the source and drone.

Hah. Good luck in pulling that off.

Bath the drone in laser pulses that blow the carrier signal into oblivion.

Signals without proper encryption are ignored. In order to perform denial of service on the drone you would have to interfere with the laser itself. To do so, you'd have to get betwen the drones and the drone carrier, at which point youve just put yourself at point-blank range AND flanked yourself. Brilliant. You're dead.

I personally want to know how you intend to ensure the drone receives this laser as it is juking away from nuclear weapons. Is the entire hull a laser receiver?

Nope, wouldn't have to be, either. Just enough surface area so that at least 80% of the time, it has a receptor surface facing the carrier. Otherwise, it can rely on standard communications systems or piggybacked orders from other drones. And, even should the drones lose contact with the carrier, they still can rely on their orders for the mission and general orders, so they're far from helpless.

This also provides excellent means for tracking the carrier, find out where the lasers are coming from ... just look for the excited hydrogen in the sea of colder hydrogen.

Sorry, come again? You want to track the path of a laser through space? Never mind that by the time you'd be close enough to try that little trick you could track the ships through the heat coming from their engines, or that it's primarily used when normal communications are being actively jammed. You really are grasping for straws here.

. Even if you could jam communications, the drones are given orders prior to launch . You couldn't jam that if you wanted to, and the drones will still be able to fire according to their mission parameters and orders.

Which goes directly back to them needing the decision capabilities as good or better than a human. It goes back to the fact that at these distances the advantages of the AI are minimal.

And human pilots aren't given orders before a mission that defines how they are supposed to act while on the mission? Get a clue.

3. Why wouldn't a ship carrying drones have ECCM to combat your ECM
Because it doesn't have the RANGE. Power goes down with r squared. If I'm jamming a drone at its distance 1 from me, and you are 10 ... you need 100 bloody times as much ECCM as I need ECM. If its 1:50 you need 25,000 times as much power.

Oh, so now we're back to knife-fighting ranges? Make up your damn mind already. First we're at long range, then we're at a range where you can slip between the drones and the carriers, and now this?

Not only that but ECCM on a carrier is asking for it to get slaughtered. ECCM acts as a giant homing beacon, something perfect for swarms of missiles to be targeted against.

Like it would be hard to make out a warship on full alert in space anyway? Whatever.

Normally carriers try to keep a low profile in actual combat, not show everyone exactly where the biggest target is.

Fine, so there are other dedicated ECM/ECCM vessels in the battlegroup. Hell, why not ECCM drones?

Not at all, and you have yet to provide any solid reasons why this would be so. With proper safeguards on the drones, the best you could do is maybe get them to shut down without alerting dozens of red flags during development, and even that would be difficult. Also, I did not ignore your scenario, I have shown how it is irrelevant, and also how it hurts you just as much with manned fighters, if not more so.

Let's say you have design team of 50 programmers with unlimited access. You have 5 people in the top brass. They decide to coup. 55 people then can bury whatever the hell they like in the code and nobody is wiser. One day the brass tells the ships to launch for training/inspection exercises and to use some set of conditions that was preset triggers the drones do ignore everything the coup leaders want it to ignore. Those fighters that are left in the hangers or not armed suffer complete AI failure. 55 people with access to numerous WMD systems, completely decimating the others.

Same damn thing can happen to your fighters, so give it a rest already.

Now what happens with pilots? The brass have to recruit pilot by pilot, these pilots aren't centralized like you programmers so its harder to bribe/threaten them all. Once you get say 50 pilots in on it you still have to deal with with the thousands of other pilots.

Not if you create a fundamental flaw in the fighter, thus rendering all the uncooperative pilots dead and their fighters free to act on behalf of the brass and the designers.

It is a bad idea to allow a few people control over all the major weapons.

First off, there's no reason to be slinging around nukes like candy even in space. Bad idea FROM THE START when a single rogue pilot can kill a few billion people with the payload on his fighter. Secondly, you know nothing of security, going by how you think crypto works and how you think a design team would be assembled. Third, the entire upper echelon has to be corrupt for this to work, in which case you're screwed anyway, and nothing will stop you from a coup.

One subsystem? I'm talking about a subtle design flaw that would cripple or destroy your fighters at any time of my choosing. You couldn't repair it without sending the fighter back to the factory to have it fixed, and that's assuming you even knew it existed. Again, you are arbitrarily assigning limitations to my position while ignoring said limitations for your own position. Brilliant.
No that is the price you pay for having a completely integrated AI ultimately in control of everything and hardcoding the whole thing.

Subtle design flaws are MUCH harder to pull off when the computer is not ultimately in charge. When individual systems have individual computers and the human can sever any connections ... it is much harder to sabotage. Think about it instead of bathering, why isn't the American nuclear arsenal controlled by computer? Why do you think that might not be a good idea here?

If you can magically subvert my sercurity protocols, I can magically subvert yours.

1. You assume that drones would be used with faulty AIs
Loads of military precedent there. The military never, never feilds equipment with bugs in it

So if it fails at all, it will always fail and be useless? Get real.

You assume that drones would be easy to subvert in the design stage
The more code the harder it is to check. Giving the AI ultimate control means it IS easier to subvert. AI's are not modular, once a peice of its code is subverted you need to check EVERYTHING ELSE. How many years does it take to completely debug gigs of code?

See above.

You assume that a rogue captain with a starship at his command is somehow more dangerous with drones than with fighters/missiles/really big cannons.
Yes because with pilots the captain needs to order the fighters to scramble, the deck crew needs to arm them, the pilots need to obey unlawful orders. Generally speaking it takes 20 people.

The only human element removed from that equation are the pilots. Deck crews and flight officers are still there, as is everyone else.

4. You assume that because drones have replaced the pilots of fighters, that they have also replaced the launch crews, maintenance teams, and commanding officers in charge of giving the drones orders.

No I assume that you have one man giving the drone orders ultimately. There is a reason why you want 2 man crews, both have to be nuts.

Now you're directly contradicting what you just said. This is getting ridiculous. And now fighters have two-man crews so that both people on board have to be nuts to do something against orders? Um, what the fuck?

1. You assume that human pilots would be able to predict one of potentially millions of possible actions of an AI, much less compensate for them.

No because an AI isn't frikking random. They don't need to plan for millions they need to figure out how the AI handles FoF and then cheat that ... or does your AI randomly open fire on friendlies?

No, but any one particular drone might not fire at the specific target you thought it would fire at. Which cup is poisoned?

You assume that fighters would come out of design without any major flaws to their structural or computer systems, and that any problems could be easily detected and fixed before they became serious.
No I'm assuming they are modular in design and that computer flaws can be worked manually in a pinch.

So I guess you missed the whole decentralized processor thing, didn't you?

The bloody fact of the matter is:
Can drones have structural flaws? Yes
Can human fighters have structural flaws? Yes
Can drones have general computer flaws? Yes
Can human fighters have general computer flaws? Yes
Can drones have AI flaws? Yes
Can human fighters have AI flaws? NO

Missed a few:
Can human fighters have human flaws? Yes.
Can drone fighters have human flaws? No.

There's some more, let's see if you can think of them.

quit dicking around with well you might have structural flaw/sabotage/whatever that is directly equivalent to an AI error. You have ALL THE SAME BLOODY PROBLEMS, plus that fact that an integrated AI ultimately holds decision power.

So what?

get into range faster than human fighters can
Irrellevant, optics gives the humans a heads up, with omnidirectional weapons the humans can engage just as easily.

Nope, sorry, you lose. Missile dance means that I can duck in far enough that you can't escape, and get out fast enough that your missiles can't catch me.

get out of range faster than human fighters can (missile dance, good against slower ships)
Useless against faster missiles. The missile is always faster than the platform.

Hello? Assuming equal missile tech, ship acceleration does become important, because slower ships can't get out of the max range, and faster ships can. Missiles can only hold so much fuel, after all, and if it's expended before if gets to me, dodging is dead easy. Your conception of space tactics is wretched.

knock down incoming missiles better than human fighters can

BS alert both sides use omnidirectional computer targeted countermeasures. The seconds gained are nothing in terms of ToF.

Say it with me, "frickin' lasers", okay? Much more accurate from a drone.

Intercept ship-killers and human fighters faster and more efficiently than human fighters can

Is directly defeated by:

have longer deployment ranges and times than human fighters
If you are father out, you have longer intercept times. Exactly how much faster is your drone?

Hello, McFly! It can use fuel more efficiently than human-piloted fighters and it can carry more of it. It is also faster than the human fighters when it needs to be. They can last longer out on patrol, though you are right that if they make advantage of their much greater acceleration they'll burn through fuel faster, but that's where the efficiency comes in.

have longer deployment ranges and times than human fighters
Yes and no. You can deploy longer/farther ... but only if it is completely autonomous or isn't jammed.

There is never a middle ground in your world, is there?

can deal with close-range combat in the event it should happen better than human fighters can.

This is drones major advantage ... and it is generally worthless.

Not really. And it's a nice bonus, not a major advantage. For a major advantage, try to understand the concept behind the missile dance.

No, really? I couldn't tell. Of course, if you've got offensive lasers, then even 300,000 kilometers is short range, but I guess you didn't consider that.
And where are going to pack your powersupply for this weapon? Isn't that in the gigawatts?

Hell, you're throwing nukes around like candy, why not some nuclear reactors too? Seriously though, we're working on lasers now based in planes capable of shooting down missiles. By the time we get to space, I would think that we could improve on the design a fair amount. Besides, how much armor can your missiles be packing, anyway, if they're so fast?

Uh-huh. And you propose to do this how? Just knowing the programming of the drone does not give you magical insight into how to control it.
Right whatever. So it is playing dice again?

Ever hear of game theory?

It's not truly random. Assume we use a seed, and you know how the computer will pick the random number. Assume that there's even a 0.00005% chance that it will pick one option more often than the others. How many battles will that win you? What use is it to you? Adding in a random factor is not worthless, according to you, but now it is? Make up your mind, or at least stop applying a double standard to your arguments.

Double standard? Whatever let's look at what you have taken for granted:
1. That your drone can identify FoF ... cannot be done now.

When you've got infantry, no, not really. However, if every friendly ship is broadcasting a secure digital signature, it's not hard at all. Civilians in war zones at in danger anyway. You wouldn't necessarily need to program in safeguards for them in particular, just a general order of "do no open fire unless fired upon" available for the drone operators to give the drones in such a situation.

2. That an equally advanced computer cannot dupe the first (remembering the first has to allow for idiot civillians).

Funny, you know, no matter how advanced a computer you have, you can't dupe mine into thinking that you're someone you're not if you can't break the encryption. It's easy one way, and potentially downright impossible the other. Thank you try again.

3. A drone that can quickly outthink a human even with millions of variables to cull through. Computers can barely beat humans at chess today ... and chess is an EASY problem for a computer ... one that has no open bounds, one where everything is in the rules.

Computers can barely beat humans at chess today? Bull. Get the best chess program you can find and crank the difficulty up to full. I'll bet you get your ass handed to you.

4. A drone that can receive orders, not be jammed or hacked, and is hardcoded.

Go back and read my original posts, it can be jammed, but it doesn't matter because it can handle enough decision making that it can still function. It can't be hacked, because the operating system is not open for the world to see, and proper encryption makes it impossible for anyone not authorized to give commands from doing so. Also, note that I was talking about a decentralized system.

Humans are not random, they are merely not as predictable. The data needed to predict humans requires far more intelligence and is far less useabl for modling. AI source code HAS to come digitized, it has to accept readily digitizable inputs, and it doesn't have things like childhood experiences clouding it. Computers have randomity or predictability. Humans are not random nor are they predictable, they appear random.

Humans are not predictable? That reeks of utter bullshit, but then I expected that.

Your random number generator is CYCLICAL. Deriving the random number values generated is not beyond the ability of computation. If you AI is random why bother with the AI, just use dice.

You clearly show zero comprehension of the points I have made, and you're repeating yourself time and again..

Even technicians would be incapable of fixing a serious design flaw. Also, if we are talking about a neural net, gee, it would figure out that its shots are going two degrees to the left of where it intended them to go. It could fix itself with that minor of a problem. That would be ridiculously easy for a neural net (or even just a normal computer program) to correct.
Really and it would know this training data how? Does it automatically go back to training mode whenever it encounters something not encoded in its banks?

I guess the concept of running maintenance between missions has never occured to you, has it? Error checking algorithms like that don't even need a neural net to work, and are ridiculously easy to code.

Neural nets take make approximations. They need to know both the input and the desired output. Eventually (assuming the problem is actually tractable) they are accurate enough to call it. However if you start presenting cases outside its ability to predict ... it needs more training data.

Right. So now it's even incapable of thrusting around properly without tons and tons of more training, right? I'm getting tired of this.

Even technicians would be incapable of fixing a serious design flaw. Also, if we are talking about a neural net, gee, it would figure out that its shots are going two degrees to the left of where it intended them to go. It could fix itself with that minor of a problem. That would be ridiculously easy for a neural net (or even just a normal computer program) to correct.
Because the enemy flies a white flag, because the enemy cuts into their frequency, because the enemy spells out we surrender in the sand.

A computer program (doesn't even have to be a neural net) can recognize these symbols and be programmed to react accordingly. Plus there's a human aspect in the Drone Controller, who can see what the drones see if he or she needs to.

First off, if drones are replacing fighters, then that means that a Drone Carrier would replace a Fighter Carrier, not a Missile Carrier. Thus, your assertion that a Drone Carrier would suck is worthless. I refer you to the points of drone superiority mentioned above.

Sigh drones can refer to UAVs, smart missiles, even drone capships. Drone fighters are the best solution using drones. You still need to trust your AI, you still need to be damn sure you can't be easily hacked and jammed. The thread starts out will their be fighters, the answer is YES. You are argueing should their be unmanned fighters, remote controlled fighters (which magically cannot be hacked or jammed), or human piloted fighters.

The answer is NO. But if you absolutely must use fighters, you're much better off using a drone-controlled fighter than a human-controlled one. I've said this since my first post.

You lose no decision-making capabilities, and you lose no significant late human control. Human-piloted fighters and bombers are still reliant on orders from their carrier before attacking. If communications are lost, they rely on their previous set of orders until contact is re-estabilished. If anything but that happens, it's usually human error on behalf of the pilots, and there is hell to pay afterwards.
Yes and they magically can't be hacked. Oh wait say encryption ten times fast and it all goes away magically.

Fine. Quantum Crypto. Poof! Decryption becomes impossible! Interception becomes impossible! Hacking becomes impossible.

And it quickly turned into "why bother using human-piloted fighters in combat". The only realistic use I can think of for human-piloted fighters is basically police duty, in-system security. Hard combat in small craft like that would be better suited to drones and drone-controlled fighters.
Hard combat is better done by smart missiles. If your fighter is really nothing more than a launching point for missiles ... which it is becoming ... then the range gets to the point where your drone has no significant advantage.

Wrong, the drone does have a significant advantage over the fighter. The fact that it's still useless in the grand scheme of things does not make your fighter any more viable.

Which, by the time we would have any sort of deep space military, I'd bet that we have the computers and programmers who could handle that easily, given the current rate of development.
Deep space military will be here as soon as space colonies exist.

Oh, yeah, that's coming soon. Forgive me if I don't hold my breath.

Meanwhile you have a gross misunderstanding of almost all the variables involved, in order to keep drones out of combat for as long as possible.
Right and which one of us thinks that a laser ranged at 100,000 km is viable?

Light has a speed of 300,000 km/s. Light can travel 100,000 km in a fraction of a second (1/3, in this case). That means that between the time a laser fires at a missile, at 100,000 km out, it has 1/3 of a second to evade that shot. Only it won't know it's being shot at until it's already been hit unless it has FTL sensors. With no atmosphere in space to diffuse the beam like on Earth, the effective range is dramatically increased to the point where you could have a laser with ranges of over a light-second.

Face the facts neural nets are not the be all end all of programming there are numerous simple things they can't deal with. You have ZIP for a garuntee they will ever be combat worthy. They introduce the possibility for hacking and jamming to screw them over more royally than human control.

Just ignore crypto and it will go away, huh? Right...

Their major advantage is they are faster than a human fighter, but slower than a missile.

Missile Dance.

With omnidirectional weapons, long time of flights, and thermonuclear weapons I'm not seeing huge savings here, what an aditional few g's of thrust and maybe 200 kg in mass savings? All while picking up increased room for error in the code, increased hacking/jamming potential, and the ability for a small number of people to take over all the fighters.

Repeating yourself, bullshit said twice is still bullshit.

I'd say it's safe to assume that there would be a "master" neural net which has all of the essential basics already figured out which could be imprinted into the new neural nets. From there, very little processing time would be needed in comparison. In any event, I don't think that you'd even need a neural net for each drone, really, but that's just me.

I think a neural net for each drone is stupid, just do one and stamp ... it works.

Okay, so we...agree?

How is that a problem, exactly? War isn't supposed to be fair , man.
[/b]
Because with long range laser weapons you need HUGE power supplies. In which case your ships will be as slow as whales in any event.

So what? If they can't evade missiles anyway, why bother? Just knock 'em down before they can even get close.

How much does a human life cost? Planning on making a clone army anytime soon?
War is a bloody mess, it will remain so. Saving pontential hundreds of lives whilst risking more with hijackable drones? I think not.

Prove they would be hijackable first.

Wow, yet another arbitrary limitation for drones. You just don't quit, do you?
No that is real world price comparison. Did you not notice the direct comparison between the predator and the F-22 ... that IS the only reason real world drones are cheaper they are piss slow, with worse weapons, and fly at low altitudes.

Actually, I'd say that it's primarily that we don't have good method for remote drones just yet. Doesn't mean we're not working on it.

There are other factors to consider here. For example:
How long does it take to make a Predator?
How long does it take to train a pilot?
How long does it take to train the operator, he he is to perform as well as a pilot he needs as much training (more because the stupid ones will survive their mistakes more frequently).

The operator would take not much longer to train than a normal flight controller. Costs are kept roughly the same.

Think more to get to the same result? Funny, it removes a ton of variables that can cloud human judgement and somehow that's a disadvantage?
It adds a ton more it has to discount in overhead processing. Computers can barely beat top chess players and chess is OPTIMIZED for computers ... its a damn simple calculation. It will require exponentially more computation to expand your dataset to battlefeild conditions.

Repeating yourself again. Repeating yourself again.

A modern day jet fighter can already handle more Gees than the pilot can. You wouldn't really need to soup up the engines or the frame that much in the first place.
And a the tolerance of a modern fighter pilot for g's keeps going up. The highest g load is still up around 80 transient.

It will never be able to keep up with the machine itself, which was the point.

Point-defense solid-state chemical lasers with computer targeting
Complete knowledge of the guidance systems of all known missiles, and their weaknesses (hey, what goes around comes around, right?), not to mention the ability to store more countermeasures than a human-piloted fighter craft and better judgement as to when to use them for maximum effect (and the reaction time to do so).
Yes modern human piloted craft don't ever use computer targetting.

The don't use computer firing, or computer manuevering, all of which can positively or negatively affect the outcome of an action.

There is didly squat a drone can do that a human with proper targeting computers can't do.

Uh-huh, right. Refer to above bit on stable firing platforms, please.

You do not need human pilots for the drones. You are limited by your military enlistment. Drones are limited by industrial capacity. Also, if drones are less likely to be destroyed, that means that during the course of a campaign, you will be outnumbered consistantly more and more as things continue. Maybe not 100 to 1, but enough so that it makes a difference. Also, remember that a drone carrier does not have to allow for pilots or pilot comforts in designing the ship, so a Drone Carrier the same size as one of your Fighter Carriers will be capable of holding many more fighters and drones than you.

Look folks it is the bait and switch. First you had the drone carrier that had humans flying these things remotely, drone control officers. Now you have a carrier with no pilots (a pitfully small number of crew on any carrier ... maybe 100 out of thousands) that doesn't need space for the pilots ... wahoo you can what 3 heads and 25 private quarters from the ship? That adds room for 5% more fighter and mass for a fraction of it? Those drone operators add nothing to ship's compliment, the fact that the mass of crew quarters and support are not likely to MASS anywhere near a single fighter is completely irrelevant.

What are you talking about? A drone operator is NOT A PILOT. Not once did I ever say that a drone operator was a pilot. I said they played the part of a flight controller. They give orders to the drones as COMMANDING OFFICERS would give orders to pilots. Jesus, no wonder you're so confused, you can't even get any of the posts straight.

Well, if you have planetary computers capable of easily decoding all the possible decisions a drone could make...

Probably not very long at all, and it would probably be pretty cheap, too.

Neural nets work because they RETURN THE SAME SOLUTION. If you have a neural net learn the function f(x,y,z,a...) and you have another one learn the function f(x,y,z,a...) and you ask each one to find a minima with tolerance of +/- delta they should ALL RETURN THE SAME ANSWERS. In order to make them react in a significantly different manner you need to have completely different functions governing their behavior ... which means you need to train each one individually with different data sets.

Further you don't need to know everything the drone will do in every circumstances, only those that it will do in a given scenario you want to know about.

First off, I'm not the one who suggested neural nets. I just said AI. Secondly, refer to game theory, and then try to comprehend the golden rule of crypto.

Hence the drone operators. There would obviously be other safeguards in place, but I guess you have assumed that everyone proposing drones has the average intelligence of a Starfleet Officer.
Drone operators lack PHYSICAL ACCESS to the systems. Pilots don't. You put a small box next to the pilot with a sequence of circuit breakers. If the computer goes to hell the pilot pulls the fuse for the system effected. If a drone operator senses his AI's are going loopy he can only send EM pulses and pray the drone receives them and isn't too far gone to act on it.

Or he lets another officer know his status and the drones IFFs are removed from the accepted list and marked as unknowns or hostiles, depending on the situation.

Even if it is more expensive than a fighter, it has distinct advantages that make it much better. Much higher efficiency, greater returns, and ZERO loss of human life should it be destroyed. War isn't about dying for your country, it's about making some other poor bastard die for his.
It also has the added headaches of more code to potentially have errors, more code to be subverted, and the hijacking/jamming problem.

Pete, repete. Pete, repete.

Zero loss of life is a good thing, but zealous pursuit of it is a good way to handicap yourself in war.

Zealous persuit to protect the human ego is an even better way to handicap yourself.

The decisions a pilot in a fighter craft makes are extremely limited. The biggest leap would be making a computer program capable of flying a fighter with some degree of skill. Not surprisingly, this is actually easier to do in space, as there are far fewer variables to have to worry about than in atmospheric flight. Beyond that, the decision-making capabilities of a pilot are marginal at best. Whether it's drones or remotely piloted fighters, the military is slowly moving away from having humans in the cockpits of fighters. Drones are simply better because they don't become entirely worthless if communications are lost, while remotely controlled fighters aren't worth a damn without constant two-way communications.
Technically speaking drones are RC fighters. All modern UAV's are drones, and are remote controlled.

No, not really. Drones are capable of making their own decisions based on their orders.

I seriously doubt humans will move out of the cockpit anytime soon. The added headaches of cyberwarfare, computer bugs, coupled with ever increasing missile mass, size, speed, and engagement ranges will make the fighter nothing more than a point to launch missiles from.

Actually, I doubt the fighter will survive at all.

Really I have yet to see an argument you have put forward that supports human occupied carriers and not just droning out the

[Hotfoot]

aw hell, I broke the character limit....

Well, at least it was nearly the last bit anyway. *reposts missing bit*

Really I have yet to see an argument you have put forward that supports human occupied carriers and not just droning out the entire force.

Bigger ships are slow and unmanueverable anyway. To get them to move fast enough to be too fast for humans and fast enough to take advantage of being drone-controlled, you'd have to put some simply massive engines on it, and it would be a terrible waste of resources.

Finally, I am ending this particular conversation here and now. I will not respond to another 6000+ word post that could easily be shaved down to a third of that size or less.

[data_link]

Notice also that this asshole's entire argument is predicated on the fact that you can hijack a drone, but you can't hijack a human pilot. Well I've got news for you asshole: if you feed a human pilot false data you can trick him into killing innocent civilians too. The formula is very simple: 1 voice vocoder tuned to mimick the commanding officer + one bullshit transmission about so and so being a traitor = several hundred human pilots attacking the wrong target. Try to do that with a drone on the other hand, you get:

"Invalid target designation. Input security verification code within fifteen seconds or prepare to be destroyed."

Thus not only saving innocent lives but also convincing the enemy that attempting to hack your drones is a very bad idea. Also, you seem to have some stupid fantasies about every program being hackable. You forget that changing the program on a properly consructed drone would entail physically ripping out the ROM and replacing it with your own, and if you can do that, you can also wire your own AI into the control circuits of a fighter that is normally human-piloted, setting it up to automatically eject the fool and then blast everyone else with far superior performance than the fighters still carrying humans because it no longer has to worry about the survival of the human pilot. Further, human pilots can be manipulated by dishonest persons to do things that they would never ordinarily do of their own volition. The fact is that it is easier to fuck up a human-controlled fighter than it is to fuck up an AI.

Also, you seem to have fantasies of people being able to figure out the operating procedures of the drones based solely on their programming. Sorry, but even today people are creating programs so complex that they dont't fully know what they are going to do until they run them. You really expect me to believe that by the time we have space fighters, we are going to have simpler computer programs? BULLSHIT. You ignore the fact that even now we have computers that rival the human brain in power... twenty years from now your desktop computer will be smarter than you. By the time we have space fighters, computer intelligence will be so advanced that it will be easy for computers to accurately simulate exactly what a human pilot is going to do and compute a perfect response to it. Simply put, humans are going to be obsolete.

Also, as to your argument about one tiny flaw fucking up the system beyond usability: that happens in humans as well, or perhaps you have never heard of cancer?

[SirNitram]

I see Tharkun is one of the idiots who think a nuclear explosion in space is large. How nice.

Here's a hint, jackass. Most of a nuclear munitions blast radius is the fact it's in atmosphere. An explosion in hard vacuum will be comparitively tiny. Even a lumbering battleship at 3G's can dodge that at long range. The intense radiation released is an issue, but a suitably large ship, or a drone, can deal with this. Unless your fighter is getting even heavier(Thus increasing it's fuel requirements, target profile, etc), it's just another downside to manned fighters.

That's another myth you seem to have fallen into, that weight doesn't matter. It still costs more to accelerate 3 tons than 1 ton.

Then we have speed. Your human piloted fighters have the same acceleration, after sufficient advancement, as a battleship. Exactly as fast as the huge battlewagon that can see farther, shoot more, and take more fire from nonnuclear ordinance. A drone, on the other hand, is limited only by the power of it's engine.

Of course, I still think neither will survive long-term. A carrier of either sort would be unnecessarily 'noisy', thus attracting stealthy capital ships to it's scent like a wounded fish draws a shark. The main reason the battleship vanished here on the ground is range limitations. In space, the only range limitation is range of detection.

[tharkûn]

1. You can out run missile's fight envelope if you can take the acceleration. (and humans can't take it)
Repeat after me the launch platform is NEVER faster than the missile. Let's say you have a drone fighter with X thrust. Now it carries 10 missiles with mass Y each and it's empty weight is 2Y. If we place an engine half as powerful on each missile they will run 10 bloody times faster. I don't know where this stupid idea came from but engines are better for small missiles than for ships carry lots of said small missiles. You do not have the speed to outrun the missiles.

2. You can still dodge lasers and rail guns what not
Rail guns are moronic. They are useful only point blank or as strategic bombardment of a hardass target. Using them anti-ship is a joke.

Lasers are a brain bug. A USMC laser target marker is less than a cm wide when it leaves the emitter, at 8 km it is *152* cm wide. Long range lasers are a bitch, anything more than a 1000km is going to be hideously inefficient. Yes lasers get wider more slowly than normal light, they still bleed out and at long ranges they are not effective weapons.

Better strategic mobility due to higher acceleration tolerance
Sigh you want to move faster? You need more propellent. More propellant means more mass which makes the process self-defeating. For a given velocity, a drone requires the SAME amount of fuel as a human fighter (give or take a percent for mass differences). Strategic mobility is determined by how far you can go before refueling ... and it isn't much further.

You are assuming nukes are going to render passive defense useless, which is not going to be the case as a bit of extra jinking means bigger warhead will be needed to ensure a kill and vice versa. You don't give away any advantage when you can have it.
Its a bitch to juke modern AtA missiles, this is why newer fighters, like the F-22, push into the ceiling. At high altitudes you can evade missile fire because the gravity well has been working against the missile. With no gravity well you are not going to be able to get significantly far from the warhead. Do you not understand that most highspeed course changes in the atmosphere are a result of friction ... not changing the thrust of the aircraft?

It is an advantage to sustain greater accelerations, but not a hideously overwhelming one. Not one I'd be willing take until the AI has proven itself more than equal to the task.

Oh a torpedo is faster than a SSN....oops it is not
1. A missile is not a torpedo, dumbass.
2. Top speed for a Shkval torpedo is around 375 kph, tell me where do I find a SSN that goes 375 kph, what class would that be again?

Humans betray, computers don't
Right and your drone operators can't do that

Humans have other motives, computers don't (causing over stating kills and such)
See above

Computers can be reprogramed faster than re-educated
Depends on the glitch. Debugging software blows, something minor that the enemy is exploiting can a bitch to find, there are bugs with windows that I still can't get MS tech support to tell me what the hell is wrong. Once you find the problem it is easier fix.

Try hacking into quantum communication. Oops you CAN'T BY THE LAWS OF PHYSICS. No, you can't hijack or intercept quantum communication with any physics we know, and having a quantum decoder would not help.
Hello dumbass you can ALWAYS hack the system, the simplest way is to simply sit at the terminal wired for the output. Inflitration of the drone ops centre if nothing else allows the drone to be comandeered. Quantum communication is HIGHLY THEORETICAL. Each and every bloody fighter needs an entangled particle which does not interact with anything else on the fighter. Good bloody luck with that.

Preferable != always
And sometimes you need big guns spinal to defeated armored missiles. (big for a small drone that is)
Big guns are moronic. Fighters have switched to missiles because guns SUCK. If it is a big missile use a AMM. At the range you need to engage a missile it can easily juke a mass driver, lasers are stretching their range to be effective and burn through.

I can program your nukes in missiles to blow too. :rolleyes: That is like saying guns are useless because gun powder blows when you burn them.
Oh get off it moron. Drones have every weakness human fighters have. Their missiles can programmed to blow, their structural systems can have flaws. Their operators can be corrupted and bribed just like pilots. In ADDITION to all this have AI weaknesses. As you continiously state the only element removed is the pilot. So every single bloody weakness a conventional fighter has, barring g limits and reaction time, a DRONE HAS TOO.

You might as well say you can prevent ICBM lauches by placing a fifty ton steel block on the launch hatch. That notion is absurd because if would be magnitudes easier to blow up the drone than blocking the comm that might be relyed from any other friendly ship in the area from any angle anywhere. Also the laser can be widened to ensure reception. As for interception, I can rely the signal so that your ships are not anywhere near the path the beam passes though, and quantum communication deals with that anyway.

Quantum communications are not a panacea, they bring their own problems ... namely the effects of general relativity on the entangled particle.

Widening the laser is a good idea, of course it makes you easier to track. Bouncing the signal is an option, but its range is limited to how much time delay you care to have and how quickly can I track the signal back with optics.

and do you even know how ECCM works?
Generally speaking there are two approaches to Electronic Counter-Counter Measures. The first is to "burn through", simply crank up the power of your transmission till it is greater than that of your enemy.

The other is to use pulse sequence (possibly multiple bands) and attempt to Fourier the jamming noise out.

Both of these have their limitations, and both can be countered. The first is critically hampered by distance, the second just requires smarter jamming. Both tactics have been dealt with extensively by modern militaries and the winner is normally the guy the most money behind his ECM or ECCM respectively.

Oh no, the designers of the ICBM lanuch systems decides to coup and attacks with a newly constructed base while disabling all others!

You do know that the US military specifically requires two men to activate ICBM launch for the very reason that the computer systems can be compromised. In the REAL world military planners realize that there is no perfectly secure computer system.

My ass, that is the problem with human organization not technological one. Besides, even with manned figthers I can just program to shut down the fly by wire system or engines with the same result.

Except that same bloody result is possible with drones. Plus the fact that if you kill the engines the pilots can just disconnect the circuit breaker and run it manually.

Who says I'll only have one team of programmers? I'll probably have dozens and if any is compromised I'll just switch.

Keep telling yourself this. Keep ignoring that real militaries DON'T use these systems, instead they use multiple people to control their system because computer security is NEVER a given.

Yes, because a spitfire has a human in it, it can dive and not suffer from fuel cut off in 1940. Anyway, remote control is no less controlled than manned ship, unless you systems fails to the point that the pilot have to physical shoot the computer to pieces.
Remote control is vunerable to jamming and hijacking. Regardless of what fanatasy you want to exist there are multiple generations of ECM and ECCM because both sides can do it. Quantum is a nice theory, of course the effects of accelerating one of the particles can blow the entanglement.

And it isn't? Only the command structure is manned, the actual weapon itself are all automatic.
Yes humans are in physical proximity to the point of launch. This is a GOOD THING because there is no substitute for physical access.

fdisk C:
format C:
xcopy D:\ C:\

cute, and useless. You presume to have a completely debugged copy in D.

And that is why you want 500 man commanding a SSBN. Opps you don't.
This is why it takes dozens of people to fire the boomers. The crew is obligated not to obey illeagal orders. The XO has the responsibility to question (privately if at all possible) any and all orders the CO makes that are illeagal.

or does your pilots randomly open fire on friendlies?
Pilots do not perfectly follow FoF protocols. You cannot bet on them not tagging you as hostile even if you are just barely fitting the class of a civillian friendly.

There is only a finite number of logical actions in any given sistuation, and given decent programing, a drone can attempt and of those with complete unpredicability. Inlogical actions should be attempted by only idiots like stupid pilots.

Which is why knowing the code works. You know which choices from which the computer will choose. You can do the same thing with humans, but you don't know what they have learned and haven't tried. If it ain't in the code/dataset, the computer WON'T DO IT.

Nuke pumped X-ray lasers your ASS.
Sorry not long enough range. Yes your missile can use a nuke pumped x-ray laser, it will not however go 300,000 km and still be a ship killer. Starwars calculations show pretty convincingly that you are not going to have a tight enough laser with enough energy.

I can and your fighter can't without computerized IFF system.
Right the US had no FoF protocols for the gulf war, vietnam, or WWII ... right

The fact of the matter is computer FoF can be fooled into thinking military craft are civillian. So long as you fit the friend classification you will always be classed it, humans are not so reliable. They have been known to hit suspiscious looking civillians on numerous occasions. Banking on the odds is not a good way to conduct war.

It beat the best human there was at the time. If beat ALL humans. Unless your pilots are all the top 0.0001% of the human population in terms of skill, you don't even have a chance at all. And chess is far more strategic while less time intensive than tactial combat, so it actually favors humans. Try playing checkers with the computer, or worst, quake 3 on hardest.
Bull. Chess is an ALGORITHM, I refer you to . Set up a function to rank board positions, plug and chug ... that and a bit of weeding software is all Deep Blue is. As you increase the number of variables the computer begins to fail, indeed the computation can become intractable. The real work is not the missile duel, but when to engage, when not to engage. All of the battle theatrics of space warfare will be done by missiles with big engines, smart tracking systems and large explosions.

MY ASS. Chess is a human game, not a computer one. If you want a game that is optimized for computers it would be a RTS running at 10,000,000 times the speed of what we normally play. (w00t lose to 6 ling in 1 micro second)

Chess is an algorithm. It is COMPLETELY computeable and quickly converges. Again I suggest you read Turing.


why fighters? why not just make long range missiles and call it a military?
Because long range missiles need a platform to fire from or have HIDEOUS ToF. If you fight beyond the optimal range, the trade off of ToF/detection/countermeasures/etc. you get burned, badly. A short range missile is MUCH faster and more maneveourable than a long range missile so if you dump a long range missile from Earth to mars it can easily be taken out by a short range missile. "Short range" (likely thousands of km) allows you to fire and not have to worry about a near infinite number of point defense shots. In a swarm you can launch sufficient missiles to overwhelm the defenses, the longer your ToF, the more rounds of point defense shots the enemy can take.

Further when the rubber meets the road (why your offense hits his defense) you want to be as fast/maneverouable as possible. A long range missile carries the fuel for the entire flight itself, it has higher overhead. A short range missile carries less fuel and has less overhead.

Basically the idea is to divide your offense into three layers:

The logistical:
The carrier. This allows you all the economics of scale (like a watch system, spair parts, repair facilities, etc.) and the best force projection. Optimally it should never see the enemy, nor should the enemy see it.

The strategic:
The fighter. This is the cheapest firing platform, it should not directly enter combat, it's only goal is to fire missiles and secondly to survive. It should see the enemy, and be seen by the enemy ... but never meet the enemy.

The tactical:
The missile. This is the where the rubber meets the road. It should be optimized for short term performance and be used to evade defenses and kill the target. It is the only element that should physically meet the enemy.

Bah these posts are too long, how about actually taking a go at the list of questions I asked?

SN:

"Most of a nuclear munitions blast radius is the fact it's in atmosphere. An explosion in hard vacuum will be comparitively tiny. "
Which is why you go for either radiation enhanced weapons or kinetic transfer weapons. Both of which were under active study by the US and Soviet military. High speed shrapnel from a nuclear blast is a good shot for an area effect kill.

That's another myth you seem to have fallen into, that weight doesn't matter. It still costs more to accelerate 3 tons than 1 ton.
I see you are an idiot. Weight is irrelevant (it's ZERO, weight measures the force of gravity), MASS is not. This is why I SPECIFICALLY ASKED how much MASS you expected the drone to save. Making half assed comments is of course superior to actually trying to give some reasonable numbers.

Then we have speed. Your human piloted fighters have the same acceleration, after sufficient advancement, as a battleship. Exactly as fast as the huge battlewagon that can see farther, shoot more, and take more fire from nonnuclear ordinance. A drone, on the other hand, is limited only by the power of it's engine.

A drone fighter will still have PISS POOR acceleration compared to a short range missile. The missile is ALWAYS faster than the launching platform. If your drone is carrying 10 missiles the the missiles are likely 10 TIMES as fast ... if not more as their engines are MORE EFFICIENT.

Of course, I still think neither will survive long-term. A carrier of either sort would be unnecessarily 'noisy', thus attracting stealthy capital ships to it's scent like a wounded fish draws a shark. The main reason the battleship vanished here on the ground is range limitations. In space, the only range limitation is range of detection.
And your operation radius and your tactical radius. The carrier offers the optimum maximum for both. Short range missiles kick the crap out of long range missiles when the rubber meets the road.

Notice also that this asshole's entire argument is predicated on the fact that you can hijack a drone, but you can't hijack a human pilot. Well I've got news for you asshole: if you feed a human pilot false data you can trick him into killing innocent civilians too. The formula is very simple: 1 voice vocoder tuned to mimick the commanding officer + one bullshit transmission about so and so being a traitor = several hundred human pilots attacking the wrong target. Try to do that with a drone on the other hand, you get:
Right yet modern drones have already been duped numerous times. Tell me, wise one, why REAL WORLD militaries keep men in missile silos? The whole works could be automated directly to the Kremlin/Whitehouse? You don't suppose the military investigated the possibility of computer control and deemed it more dangerous than manual control?

Also, you seem to have fantasies of people being able to figure out the operating procedures of the drones based solely on their programming. Sorry, but even today people are creating programs so complex that they dont't fully know what they are going to do until they run them.
Learn to READ. I said to take the source code and RUN it MILLIONS of times on a super computer. Did I EVER say you could glance at the code and defeat the drone of death? No. I said you take it and run it on a computer with TWO ORDERS OF MAGNITUDE faster processor.

By the time we have space fighters, computer intelligence will be so advanced that it will be easy for computers to accurately simulate exactly what a human pilot is going to do and compute a perfect response to it.
Right and this is based on what exactly?



So come on boys are you to scared to answer the relatively simple questions I asked, or do plan to continue attacking strawmen and patting yourselves on the back?

[SirNitram]

None of us are utilizing strawman arguments. We are dismantling your argument with real logic and real science. You, on the other hand, are ignoring the benefits of drones and just mindlessly repeating yourself and declaring victory.

[tharkûn]

None of us are utilizing strawman arguments. We are dismantling your argument with real logic and real science. You, on the other hand, are ignoring the benefits of drones and just mindlessly repeating yourself and declaring victory.
BS. You talk about weight mattering ... it doesn't ... MASS does. Something I explicitly asked about yet you are too scared to answer the question.

10. How much mass/space (profile only) will you save going drone?

Come in it's not that hard.

Your grasp of science has not been demostrated. Force = MASS x acceleration. WEIGHT = m * g. So in your world does force have units of kgm^2/s^4? Oh wait we wanted REAL science

Data_link EXPLICITLY attacked a strawman.
Also, you seem to have fantasies of people being able to figure out the operating procedures of the drones based solely on their programming. Sorry, but even today people are creating programs so complex that they dont't fully know what they are going to do until they run them.
I NEVER suggested you could figure out what a program does without running it. I simply suggested you RUN IT A MILLION TIMES.

[SirNitram]

Wow, you can nitpick the difference between weight and mass... You must be so proud.

You want to know how much you save? No cockpit. No controls. No circuits to move the info along. No life support. No windows. No supplies for the pilot.

You want to know about speed? A human fighter is restricted to 30G, and that's bursts. Again, anything you can squeeze out of an engine is your speed for drones.

Are you going to refute either of these, idiotboy?

[tharkûn]

Wow, you can nitpick the difference between weight and mass... You must be so proud.
Yes REAL SCIENCE doesn't care about the difference between weight and mass

But hey when you blow it, just be sure to say the other guy is "nitpicking" it distracts from your failure.


You want to know how much you save? No cockpit. No controls. No circuits to move the info along. No life support. No windows. No supplies for the pilot.
Gee at most 1 m^3. Of course the drone requires space for its expanded computer system (which is going to be big since everything is hard coded), cooling systems. In short you save enough for a few extra litres of fuel ... whoop de do.

You want to know about speed? A human fighter is restricted to 30G, and that's bursts. Again, anything you can squeeze out of an engine is your speed for drones.
Yes I want to know about speed. Speed is the norm of velocity, it has units of m/s. G's are a unit of acceleration (m/s/s) ... not speed. Tell me do you help B&B with their unit consistancy on Star Trek? Does point out your abysmal grasp of science count as nitpicking here ... you did ask a question?

In any event top speed is NOT determined by your engine, it is determined by how much fuel/propellant you carry compared to your mass. With equal amounts of propellant and fuel ... they both have the SAME top speed. Yes a drone can accellerate faster than a human, however this may well be meaningless. This is why I ask for this:
"3. How much faster will a drone be (i.e. how many m/s/s will a human fly, how many will a drone)? How much more fuel will it carry for these speeds (faster engines being less efficient than slow ones normally)? "


The fact of the matter is, no matter how much faster your drone is, a short range missile will be faster BY FAR. Optics will kick the crap out of both. So you end up with a missile duel, in which maneuvoring is of limited (if any) value.

[SirNitram]

Oh no, Tharkun is being anal about units when the meaning is clear. My argument is so falling to peices...

Once you're done, you will note I've only used units of measurement that cooincide with acceleration and mass. But things like facts never stop you.

As to your ignorant claims of just 1m^3? I'd love to see a reference. Keep in mind you will need more than a modern fighter jet has, because.. surprise! we're in hardvac. And I'm sure you'll be equally anal as always and try and paint me as some sort of troll because I don't spell out 'Hard Vacuum'.

Wow, you figure out basic physics too.. Never would have thought you had it. But a drone has less mass, so it requires less fuel, so it costs less. And, of course, you completely ignore the issue of Human G-Tolerance issues. Because, of course, addressing it would be realizing you're blatantly, totally, wrong.

As for missiles? Of course they are. Hence why I advocate heavy cruisers, submarine style, with large missile armanant. I'm just dispelling your ignorant notion human-piloted fighters are feasible against drones.

[Hotfoot]

Repeat after me the launch platform is NEVER faster than the missile. Let's say you have a drone fighter with X thrust. Now it carries 10 missiles with mass Y each and it's empty weight is 2Y. If we place an engine half as powerful on each missile they will run 10 bloody times faster. I don't know where this stupid idea came from but engines are better for small missiles than for ships carry lots of said small missiles. You do not have the speed to outrun the missiles.

Allow me to paint you a picture. You have two platforms capable of firing the same missiles. One platform is faster than the other. They start off outside of each other's maximum missile range and start off by closing the gap so that they can fire missiles. Now, the thing you have to keep in mind that the maximum range of the missile is limited to the fuel it can carry for thrust. Beyond that, it becomes a dumbfire weapon and cannot adjust its course, at which point changing course just a little will let even the largest capital ship avoid being hit. The obvious point here is that you don't fire the missile at your target the second your target gets in range, as it can change velocities and head out of range of the missile before it hits. It doesn't matter how fast the missile moves if the target can get force it to expend all of its fuel before it can kill the target. Thus, you have to wait until the target is too far inside the reach of the missile to be able to pull out, and then fire, removing the option of a missile dance.

Here's a rough equation you can use to figure this optimal firing distance out yourself.

b = burn time (how long missile can accelerate)
a = acceleration rate of missile (m/s/s)
v = starting relative velocity to target (m/s)
c = target's acceleration rate (m/s/s)
d = initial distance between missile and target
D = max range of missile

We'll assume that the target has enough fuel that it will not run out in this encounter.

Maximum range of the missile is D = v * b + (1/2a * (b*b))

Now, for the sake of simplicity, we'll assume the target is stationary and subtract its acceleration rate from that of the missile.

The equation for determining if the missile can catch the target is:

d < v * b + (1/2(a-c) * (b*b))

What does this mean? Well, let's plug a few somewhat arbitrary numbers in there, shall we? Lets assume that you missile can go 300m/s/s with a burn of 30 seconds and starts off with a relative velocity of 400m/s to the target. The target is a human-piloted fighter, and the best acceleration it can pull is 90m/s/s. Maximum range of the missile is 147,000 m, or 147km. In order to ensure that the human-piloted fighter cannot evade the missile by outrunning it, you have to close to at least 106.5 km.

Alternatively, say you have a drone-controlled fighter with just twice the acceleration rate of the human-piloted fighter, 180m/s/s. In order to ensure that the drone-controlled fighter, you must close to at least 66km.

What does this mean? It means that the drone fighter has roughly a 62% range advantage over the human fighter. The drone fighter can simply dive in to 106 km, fire off a missile that the human pilot can't outrun, and be well outside of the maximum range of the human fighter's missile the whole time. In other words, faster missile platforms have longer ranges than slower missile platforms using the same missiles.

And that is how you can outrun a missile.

Rail guns are moronic. They are useful only point blank or as strategic bombardment of a hardass target. Using them anti-ship is a joke.

Depends on how fast you can accelerate the projectile.

Lasers are a brain bug. A USMC laser target marker is less than a cm wide when it leaves the emitter, at 8 km it is *152* cm wide. Long range lasers are a bitch, anything more than a 1000km is going to be hideously inefficient. Yes lasers get wider more slowly than normal light, they still bleed out and at long ranges they are not effective weapons.

Lasers are a brain bug? Right. On Earth, lasers disperse more than they would in space thanks to our atmosphere. Dispersion of a laser in space is very minimal. Dispersion of a laser is much, much slower than that of unfocused light. Now, I will direct you to the military laser currently under development which can track, target, and destroy a super-sonic artillery shell, in the Earth's atmosphere. I also ask you again, just how much armor do you plan on putting on your missiles in the first place? The better that they can withstand a point-defense laser, the more time the other guy is going to have to shoot at it.

<snip>
Hello dumbass you can ALWAYS hack the system, the simplest way is to simply sit at the terminal wired for the output. Inflitration of the drone ops centre if nothing else allows the drone to be comandeered. Quantum communication is HIGHLY THEORETICAL. Each and every bloody fighter needs an entangled particle which does not interact with anything else on the fighter. Good bloody luck with that.

Quantum communication is NOT highly theoretical, we can do it today, and it's much farther advanced than quantum computing. We can even transmit quantum encryption over the air for a short distance. Do you even know what you're talking about? Quantum Encryption in particular is incredibly easy to do, as it uses polarization of light to transmit the key. You cannot intercept the key, you cannot break the encryption, you cannot hack the system. The best you can possibly hope to do is denial of service when Alice and Bob are attempting to transmit the key, and that's impossible if the key is established on the ship. You lose.

Big guns are moronic. Fighters have switched to missiles because guns SUCK. If it is a big missile use a AMM. At the range you need to engage a missile it can easily juke a mass driver, lasers are stretching their range to be effective and burn through.

The longer the range of the missile in space, the more time you have to shoot it down or evade.

<snip more useless repetition>
Which is why knowing the code works. You know which choices from which the computer will choose. You can do the same thing with humans, but you don't know what they have learned and haven't tried. If it ain't in the code/dataset, the computer WON'T DO IT.

So, because humans are limited to fewer options than drones, they're...less predictable? That's bullshit.

Sorry not long enough range. Yes your missile can use a nuke pumped x-ray laser, it will not however go 300,000 km and still be a ship killer. Starwars calculations show pretty convincingly that you are not going to have a tight enough laser with enough energy.

So you're putting as much armor on each of your missiles as your capital ships? Wow, those missiles must have shit for acceleration rates then...

<snip more repetition>
Bull. Chess is an ALGORITHM, I refer you to . Set up a function to rank board positions, plug and chug ... that and a bit of weeding software is all Deep Blue is. As you increase the number of variables the computer begins to fail, indeed the computation can become intractable.

Oh goodness no. Chess is not an algorithm. Chess is a game. An algorithm is a step-by-step problem solving procedure. Pick up a dictionary and get a clue.

How a computer goes about playing chess is an algorithm. Chess is the problem to be solved. Guess what, humans use algorithms to play chess as well. Well, at least the good ones do. In fact, everything we do can be described as an algorithm, from walking to eating to having sex to giving birth to playing soccer.

The real work is not the missile duel, but when to engage, when not to engage. All of the battle theatrics of space warfare will be done by missiles with big engines, smart tracking systems and large explosions.

See above equations. Also, how the hell are you going to get a large explosion in space? The absence of an atmosphere really screws that whole idea up. If you want an area-effect weapon, you're probably better off making some sort of fragmentation missile that spews superheated kinetics all over the place at high velocities. Of course, since it's an unfocused blast, most of those will likely wind up missing the target.

Chess is an algorithm. It is COMPLETELY computeable and quickly converges. Again I suggest you read Turing.

You've read Turing? Surprising, if you're calling chess an algorithm.

<even more snipping>
The strategic:
The fighter. This is the cheapest firing platform, it should not directly enter combat, it's only goal is to fire missiles and secondly to survive. It should see the enemy, and be seen by the enemy ... but never meet the enemy.

Drone fighters have a massive advantage over human fighters in this regard with just twice the acceleration.

<snip snip snip>
Which is why you go for either radiation enhanced weapons or kinetic transfer weapons. Both of which were under active study by the US and Soviet military. High speed shrapnel from a nuclear blast is a good shot for an area effect kill.

A directed fire weaon is so much more efficient than an unfocused shrapnel-nuke it isn't even funny.

I see you are an idiot. Weight is irrelevant (it's ZERO, weight measures the force of gravity), MASS is not. This is why I SPECIFICALLY ASKED how much MASS you expected the drone to save. Making half assed comments is of course superior to actually trying to give some reasonable numbers.

Weight can easily be converted to mass, so long as we understand that the gravitational force of Earth is the intended gravity.

<snipping again>
So come on boys are you to scared to answer the relatively simple questions I asked, or do plan to continue attacking strawmen and patting yourselves on the back?

Let's see, you don't know what an algorithm is, you don't know what quantum cryptography is, you don't know what greater acceleration rates mean to a missile platform, etc. and so on.

[data_link]

ta_link EXPLICITLY attacked a strawman.
Also, you seem to have fantasies of people being able to figure out the operating procedures of the drones based solely on their programming. Sorry, but even today people are creating programs so complex that they dont't fully know what they are going to do until they run them.
I NEVER suggested you could figure out what a program does without running it. I simply suggested you RUN IT A MILLION TIMES.

Wow... so you figure out what the program does by running it a million times. And exactly how is this different from attempting to figure out an enemy's strategy by observing his actions in battle? Oh wait... it's the fact that his opponent will not be your human pilots, but rather just your best attempt at a computer-based fighter program (which, due to your reliance on human pilots, will be substantially inferior to that of your enemy), so your simulation still tells you precisely NOTHING about how the drone behaves in real combat. Face it - your only real source of data is the same data that you would have about a human pilot's behavior, and frankly, your attempts to pretend otherwise are simply pathetic.

Also, thank you for nitpicking instead of answering the point of my post, which was that:

A. AI is less easily fooled than humans.
B. You will not be able to determine exactly how the drone will operate in combat even WITH full knowledge of it's programming.
C. Humans are subject to more flaws than AI's in general.

Repeat after me the launch platform is NEVER faster than the missile. Let's say you have a drone fighter with X thrust. Now it carries 10 missiles with mass Y each and it's empty weight is 2Y. If we place an engine half as powerful on each missile they will run 10 bloody times faster. I don't know where this stupid idea came from but engines are better for small missiles than for ships carry lots of said small missiles. You do not have the speed to outrun the missiles.

You'll be able to do a lot better at evading a missile or other weapons system if you can sustain more than 6 G's of acceleration.

Rail guns are moronic. They are useful only point blank or as strategic bombardment of a hardass target. Using them anti-ship is a joke.

Lasers are a brain bug. A USMC laser target marker is less than a cm wide when it leaves the emitter, at 8 km it is *152* cm wide. Long range lasers are a bitch, anything more than a 1000km is going to be hideously inefficient. Yes lasers get wider more slowly than normal light, they still bleed out and at long ranges they are not effective weapons.

What the hell does the effectiveness of weaponry have to do with the practicality of AI pilots?

Sigh you want to move faster? You need more propellent. More propellant means more mass which makes the process self-defeating. For a given velocity, a drone requires the SAME amount of fuel as a human fighter (give or take a percent for mass differences). Strategic mobility is determined by how far you can go before refueling ... and it isn't much further.

A percent? My dear boy, one who accuses others of building strawmen should never attempt to make straw giants.

Its a bitch to juke modern AtA missiles, this is why newer fighters, like the F-22, push into the ceiling. At high altitudes you can evade missile fire because the gravity well has been working against the missile. With no gravity well you are not going to be able to get significantly far from the warhead. Do you not understand that most highspeed course changes in the atmosphere are a result of friction ... not changing the thrust of the aircraft?

He understands that... he also understands that removing a major source of mass (life support + human) will substantially improve the performance of the fighter.

It is an advantage to sustain greater accelerations, but not a hideously overwhelming one. Not one I'd be willing take until the AI has proven itself more than equal to the task.

If you don't think AI's are equal to the task, then I dare you to play your chess program on the highest setting and win. Let me know when you realize that you can't do it (unless you are a grandmaster, and even then Deep Blue would still own you).

Right and your drone operators can't do that

Right... and your commanding officers don't do that.

Depends on the glitch. Debugging software blows, something minor that the enemy is exploiting can a bitch to find, there are bugs with windows that I still can't get MS tech support to tell me what the hell is wrong. Once you find the problem it is easier fix.

What kind of idiot lets Microsoft program their fighter AI?

Find any such bugs in a system programmed by someone who cares more about reliability than about sales. Linux, for instance.

Hello dumbass you can ALWAYS hack the system, the simplest way is to simply sit at the terminal wired for the output. Inflitration of the drone ops centre if nothing else allows the drone to be comandeered. Quantum communication is HIGHLY THEORETICAL. Each and every bloody fighter needs an entangled particle which does not interact with anything else on the fighter. Good bloody luck with that.

Oh goody, you have someone else sitting at the terminal wired for the output. And exactly how would you get there, hmm? How would you thwart the biometric identification systems that are likely to be present at any critical piece of equipment. And exactly how would this be diferent than commandeering a fighter for your own use... especially considering that unlike a fighter, where you can escape back to your home base, in the case of you stealing drones the ships' crew would know EXACTLY where you are, storm in, and either shoot you in the head or capture yoiu for interrogation and then shoot you in the head. Sorry, but all you have proven is that drones are a bad idea if your security sucks as much as it does on Star Trek, in which case you'd be fucked anyway.

Big guns are moronic. Fighters have switched to missiles because guns SUCK. If it is a big missile use a AMM. At the range you need to engage a missile it can easily juke a mass driver, lasers are stretching their range to be effective and burn through.

Um... that's why you put LENSES on your lasers, dumbass. And exactly what is the effectiveness of point-defense mechanism supposed to prove, except that human fighters have no chance whatsoever of surviving?

Oh get off it moron. Drones have every weakness human fighters have. Their missiles can programmed to blow, their structural systems can have flaws. Their operators can be corrupted and bribed just like pilots. In ADDITION to all this have AI weaknesses. As you continiously state the only element removed is the pilot. So every single bloody weakness a conventional fighter has, barring g limits and reaction time, a DRONE HAS TOO.

Get off it yourself. And enough with the strawmen. The point is that EVERY SINGLE WEAKNESS A DRONE HAS, A FIGHTER HAS TOO, ON TOP OF WHICH YOU ADD ALL THE WEAKNESSES OF A HUMAN PILOT. Do you get it now?

Quantum communications are not a panacea, they bring their own problems ... namely the effects of general relativity on the entangled particle.

Widening the laser is a good idea, of course it makes you easier to track. Bouncing the signal is an option, but its range is limited to how much time delay you care to have and how quickly can I track the signal back with optics.

This whole thing is pointless anyhow, as if you can prevent data transmission to your drones, you can also prevent sending orders to human pilots.

You do know that the US military specifically requires two men to activate ICBM launch for the very reason that the computer systems can be compromised. In the REAL world military planners realize that there is no perfectly secure computer system.

Including humans... which I guess is why they need two of them.

Except that same bloody result is possible with drones. Plus the fact that if you kill the engines the pilots can just disconnect the circuit breaker and run it manually.

Not if they're not alive. Just program the computer to open the cockpit while in space...

Besides, what kind of an idiot do you have to be to think that hotwiring the systems would allow you to achieve any kind of reasonable combat performance?

Keep telling yourself this. Keep ignoring that real militaries DON'T use these systems, instead they use multiple people to control their system because computer security is NEVER a given.

No, the reason they don't use these systems is because OPERATOR security is never a given. Computer security IS a given, otherwise there would be no such thing as a trusted system.

Remote control is vunerable to jamming and hijacking. Regardless of what fanatasy you want to exist there are multiple generations of ECM and ECCM because both sides can do it. Quantum is a nice theory, of course the effects of accelerating one of the particles can blow the entanglement.

What idiot seriously suggested that these fighters would be remote controlled? Oh wait, that's your strawman. Nevermind.

Yes humans are in physical proximity to the point of launch. This is a GOOD THING because there is no substitute for physical access.

I take it that you intend to make it possible for pilots to easily hotwire your systems then? Great... I'll never have to manufacutre my own fighters again... I'll just steal yours!

This is why it takes dozens of people to fire the boomers. The crew is obligated not to obey illeagal orders. The XO has the responsibility to question (privately if at all possible) any and all orders the CO makes that are illeagal.

And you think that drones can't be programmed to ignore illegal orders? I'd like to see just one piece of solid evidence ot support this idiotic assumption.

Pilots do not perfectly follow FoF protocols. You cannot bet on them not tagging you as hostile even if you are just barely fitting the class of a civillian friendly.

Exactly, and there are no friendly-fire casualties in modern war because humans have better judgment. Oh wait, there are.

Which is why knowing the code works. You know which choices from which the computer will choose. You can do the same thing with humans, but you don't know what they have learned and haven't tried. If it ain't in the code/dataset, the computer WON'T DO IT.

You seem to have some serious pre-concieved notioins about the flexibility of computers. Once again, I point out that you CANNOT predict what a computer program wil do with a particular input set without actually RUNNING the program, and unless you have some magical way of predicting exactly what the conditions for a given battle are, your damn simulations are useless.

By the way, his point was that there are a finite number of POSSIBLE actions that won't result in you just getting shot. Frankly, if you can tell me a comlete list of those actions in any given situation, then you can anticipate a human's actions as well (and if they do something that's not on that list, they're already dead).

Right the US had no FoF protocols for the gulf war, vietnam, or WWII ... right

The fact of the matter is computer FoF can be fooled into thinking military craft are civillian. So long as you fit the friend classification you will always be classed it, humans are not so reliable. They have been known to hit suspiscious looking civillians on numerous occasions. Banking on the odds is not a good way to conduct war.

And oftentimes that turned out to be the wrong decision. Human FoF can be fooled more easily than human friend-or-foe, because humans are not as perceptive as computers and therefore cannot check nearly as many variables. Oh, and please provide some actual data to back up your bullshit about human instinct being in any way reliable, especially given that it hasn't prevented numerous darwin awards from being distributed.

Bull. Chess is an ALGORITHM, I refer you to . Set up a function to rank board positions, plug and chug ... that and a bit of weeding software is all Deep Blue is. As you increase the number of variables the computer begins to fail, indeed the computation can become intractable. The real work is not the missile duel, but when to engage, when not to engage. All of the battle theatrics of space warfare will be done by missiles with big engines, smart tracking systems and large explosions.

So you ADMIT that computers outperform humans in the tactical aspects of battle. Concession accepted.

Chess is an algorithm. It is COMPLETELY computeable and quickly converges. Again I suggest you read Turing.

Completely computable my ASS. Do you have any idea about how long it would take our best supercomputers to compute EVERY possible chess game? Here's a hint: it's longer than the present age of the universe.

"Most of a nuclear munitions blast radius is the fact it's in atmosphere. An explosion in hard vacuum will be comparitively tiny. "
Which is why you go for either radiation enhanced weapons or kinetic transfer weapons. Both of which were under active study by the US and Soviet military. High speed shrapnel from a nuclear blast is a good shot for an area effect kill.

Radiation would kill human pilots far faster than drones.

Shrapnel is irrelevant. You will be assimilated. Oops... wrong form.

Shrapnel attached to the nuke itself would be vaporised by the blast and therefore unable to affect anything. You're thinking about shrapnel from objects destroyed by the blast and then accelerated outward by the shcokwave, which would be nonexistent since there are no such objects and there is no shockwave.

I see you are an idiot. Weight is irrelevant (it's ZERO, weight measures the force of gravity), MASS is not. This is why I SPECIFICALLY ASKED how much MASS you expected the drone to save. Making half assed comments is of course superior to actually trying to give some reasonable numbers.

Oh brother. So you can nitpick the difference between mass and weight. So fucking what? It still costs more to impart the same acceleration to objects of higher mass.

A drone fighter will still have PISS POOR acceleration compared to a short range missile. The missile is ALWAYS faster than the launching platform. If your drone is carrying 10 missiles the the missiles are likely 10 TIMES as fast ... if not more as their engines are MORE EFFICIENT.

You know, your entire argument seems to be that the superior performance of drones as a launch platform, either for missiles or direct-fire weapons, is irrelevant because the launching platform is dead anyway. That being the case, you have just proven that drones are the only option because humans would never agree to pilot those things if they are going to die anyway... concession accepted.

And your operation radius and your tactical radius. The carrier offers the optimum maximum for both. Short range missiles kick the crap out of long range missiles when the rubber meets the road.

Which is why you use fighters as the launching platform. Stop being wishy-washy and idstracting from your main argument so that we may point out how crappy it is.

Right yet modern drones have already been duped numerous times. Tell me, wise one, why REAL WORLD militaries keep men in missile silos? The whole works could be automated directly to the Kremlin/Whitehouse? You don't suppose the military investigated the possibility of computer control and deemed it more dangerous than manual control?

May I point out that the U.S. military enacted that policy when "computer" meant something that takes up an entire room? May I point out that back then, it was actually possible to break codes without knowing the key, given enough time? The U.S. created that policy back when computer security was pitiful compared to what it is now. Not only are you appealing to authority, but you are appealing to irrelevant authority, and an authority that has strong tendencies toward tradition, rather than imagining the possibilities of things in the future. Or do I need to remind you that at one time the U.S. military thought airplanes had no military value?

Learn to READ. I said to take the source code and RUN it MILLIONS of times on a super computer. Did I EVER say you could glance at the code and defeat the drone of death? No. I said you take it and run it on a computer with TWO ORDERS OF MAGNITUDE faster processor.

This assumes that you have a computer with two orders of magnitude faster processor, and that you can also find some way of determining what it will do when exposed to the actual situation, as opposed to just the input you feed it. AND it assumes that every copy of the program will be EXACTLY the same, AND it assumes that it won't be modified significantly while you're doing your analysis, AND it assumes that you can actually IMPLEMENT the proper response, AND it assumes that your enemy didn't run the AI through a similar process while designing it in order to ensure that finding a foolproof strategy against it is IMPOSSIBLE. Frankly, I'm getting tired of these bullshit assumptions.

Right and this is based on what exactly?

A simple extrapolation of the computing power of $1000 (today's money) computers over the last 100 years, as outlined in The Age of Spiritual Machines by Ray Kurzweil. Assuming one neuron firing = one computation (a dubious assumption at best), a $1000 computer will have the computing power of one human brain by 2025. Get the picture? Good. Now stop with the strawmen and let's all play a nice game of fizzbin.

[tharkûn]

Once you're done, you will note I've only used units of measurement that cooincide with acceleration and mass. But things like facts never stop you.

1 g = 9.8 m/s/s

As to your ignorant claims of just 1m^3? I'd love to see a reference. Keep in mind you will need more than a modern fighter jet has, because.. surprise! we're in hardvac. And I'm sure you'll be equally anal as always and try and paint me as some sort of troll because I don't spell out 'Hard Vacuum'.
Remember kids when you screw up, be sure to call the guy who calls you on it names ... it makes people forget how dumb you look.


As far as dimensions go ... let's say your average pilot is 1.8 (6 ft)metres tall, now of course the seated pilot will not need that much head room, but we can be generous. Now let's give him .7 m (3.2 ft) in both of the other dimensions (about enough room to fully extend his arms forward and halfway on each side). Now what about life support? Well screw cabin life support ... go with a flight suit. Water takes up minimal room, heating is done by lacing mildly resistive wires through the flight suit and plugging the unit into a jack, and air can be taken care of with nice scrubbing system (oxidizers regerated by electrochemistry). All told 1 m^3 is not that bad. Modern fighters have a canopy that allow the pilot to see, this is useless in space (when anything you can see is likely too late to react against anyways).

Of course you are the one who claims this will result in significant space savings for the drone, so feel to show better numbers.

Wow, you figure out basic physics too.. Never would have thought you had it. But a drone has less mass, so it requires less fuel, so it costs less.
How much less mass? The F-22 masses in at 18150 kg (minimum takeoff mass), of that perhaps 200 goes pilot and his support equipment. The VAST majority of the mass goes to the weapons, frame, and fuel.

In space these will require a much higher amount of the total mass.

So the drone masses 1.2% less (best case scenario for the drone) ... pardon me for not being impressed.

And, of course, you completely ignore the issue of Human G-Tolerance issues. Because, of course, addressing it would be realizing you're blatantly, totally, wrong.
Humans can't withstand high G's ... so what? You AREN'T dogfighting, you can't use air resistance for quick maneuvors. The only way to get off your trajectory is to fire the engines in a different direction. How much fuel do you intend to pack? The saturn IV went through 2 MILLION kg of fuel in 160 seconds (its empty mass was 130,000 kg), going with Newtonian physics you aren't going to get many orders of magnitude better.

Exactly how much acceleration do you plan to use? Why in hell wouldn't a simpler short range missile be able to outmanuever your bigass lumbering drone (and my bigass lumbering fighter) like a Indy car around a Modle A Ford?

What you people don't get is your velocities are limited by how much FUEL you carry. Due to the amount fuel you carry there is NO WAY IN HELL you can come close to missile acceleration or velocity. Given that you are detected by optics LONG before you are in range you will not be able to do some crappy missile dance ... you do NOT have the fuel to do multiple extended burns. Yes instead of burning for 60 seconds at some low g acceleration you can burn for 30 at something higher ... WHO CARES?

As for missiles? Of course they are. Hence why I advocate heavy cruisers, submarine style, with large missile armanant. I'm just dispelling your ignorant notion human-piloted fighters are feasible against drones.
All fighters are is a launch platform for missiles. They are inherently superior to heavy cruisers because they can attack from more vectors, if you lose they still take a good number of the enemy with them, and they present a much better target profile. Why in hell would you see a shift backwards to cruisers from fighters?

To repeat:
1. Your velocities are limited by your fuel capacity.
2. Higher thrust engines are less efficient.
3. For every kg of empty mass you need about 20 of fuel for high g burn. This means you have 5 minutes of top acceleration max.
4. Fighters/fighter-drones/longrangemissile drones all have to carry so frikking much fuel they are VASTLY slower than short range missiles.

Drone advantages are:
a. Higher g loads available.
b. Slightly more space/mass for weapons, fuel, bigger engines, whatever (but you only get to use said mass on ONE).
c. Faster reaction times.

a. Is rather useless given that you are limited by the AMOUNT OF FUEL, not the engine performance. There is no friction for high speed maneouvers, there is no point to a snap turn ... at the ranges of engagement you are looking at you will have omnidirectional weapons.
b. Is not much, at best it is 5% of total volume and 2% of total mass, noth of these are GRATUITIOUSLY GENEROUS on my part.
c. Is not an issue as ToF/optics makes split second decisions less important.

[SirNitram]

Your idiocy is staggering, Tharkun.

I see you know the meaning of a G. Congratulations.

Your overoptimistic claims on the size of life support show you truly know nothing of space flight.. Do a little research, please. Check, say, the Apollo project. Compare the craft used to the unmanned craft. But you won't, and you'll spew the same stupidity. One meter squared indeed. Of course, the possibility exists for technology to advance that much.. But by that point, computers will have miniaturized to the point a drone is an engine with missile pods.

Nice claims about high G accel being impossible. You obviously don't keep track of the real world. A Martian probe a year or three back made accel and deccel that would laminate human pilots to the underside. The trip took it only a week. But you ignore this.

Yes, fuel. I'll come back to this in your ignorant spewings later, but one need not approach the speed of hte missiles. One need only keep out of range of your opponent. Another fact I will be returning to. And, you ignorant little boy, accel matters. Despite your ignorance, nuclear weapons in vacuum do not have blasts dozens of klicks in diameter. And with proper ECM and PD, one can shoot one down.

Now, your ignorant claims about fighters being better. Allow me to expound on why you're a moron.

1. A cruiser has more space for more fuel, and is deployed for such long periods it can easily pick up speed slowly.
2. A high thrust engine is not needed.
3. That's the most ridicuosly arbitrary number I've seen you throw out. With a large vessel, one can easily ration fuel better.. But you think on such stupid terms, I can't see how you got this number and think it appears to everything.
4. Hence why they are unneeded.

You, of course, just magically assume a fighter is better than a battleship. Why? The only reason a battleship is crippled on Earth is the CURVE OF THE EARTH. Their projectile weapons are limited to near LOS. Missiles work well, but these suffer from gravity and the fact you can't coast in a gravity well. Remove the gravity well and curve of the Earth, and voila.

You insist on expensive, short range, low G, HUMAN PILOTED fighters for space. Their carrier is still a target, you know. And a fighter, of course, is a bright light moving around in space from all the high energy systems, having to expend fuel so often, etc. Compared to my theorized submarine-style cruisers, they'd never notice it until it had already unloaded it's full missile load into space heading for them. What would the fighters do? Certainly not catch up with the missiles. Intercept the cruiser itself? Keep in mind range of weapon is going to be more often than not the range of detection, and my cruisers run silent, not burning hot like a fighter swarm.

But you'll come back with more of the same bullshit... Same as always.

[Xon]

Windows 95 was a commerical OS, written with backwards compatibility for MS-DOS which was written for backwards compatibility with some other really old funky system.

Windows 95 reused alot of the code from the latest version of MS-DOS & windows 3.11, and thus was limited in what it could do. Yet it still sold like very very well. For an OS it is crap, but as a business venture it was very successful. And since Microsoft is a business, which likes to make money, ...

And you are trying to compare Win95 to a miltary program??

You start saying something and sound like you know what your talking about, then you pull something like this crap out of your ass.