[Ultra-Tech] What does the TL10 battlefield look like

[acrosome]

What the heck, I'll throw in my $0.02 on the two contentious issues being discussed:

1. Laser ground stations vs. a mobile force.

I'll go with a continuation of modern reality as being most likely in the future- with force balance being anywhere nearly equal a nonmobile fortress will always fall to a mobile force. Think Eben Emael or the Maginot Line.

That orbital defense laser ground station that is being proposed as the most bestest thing evar is going to be meat. For one thing mirrors work both ways, so the orbital ships can just bathe it in laser light and then it can't open its presumably armored shutters, or the laser dies. The ships can alter their orbits to come over the horizon at an unpredictable time from multiple directions and then turn on the lasers to bathe the ground station, yet the ground station cannot do the converse. Then, the ships launch impactors. If the ground station opens it's shutters to PD the impactors then it's at least a firepower kill, because the laser dies (and then the impactors hit anyway). And if it doesn't then the impactors attack unopposed, probably as a catastrophic kill. And it's already a mobility kill just by being a ground station. That's actually a decent point- consider that stopping a tank or ship from being able to move is, in modern combat, considered a type of "kill."

For that matter, so long as your spies are even minimally competent, the people with the spaceships are going to know where the ground station is. Forever, because it can't move. So, they can arrange a swarm of relativistic impactors to hit it from all the way across the system (assuming that's possible in the setting, but even nonrelativistic but nonetheless really fast impactors will work) and the station can't dodge! Stealth them and put ECM on them (to the extent that you can, in space) and launch enough that the laser has no chance at PDing them all before they hit. Then, you don't even need the orbiting ships. They just come along later and drop the occupation troops.

Now, to argue the other way, this is all more complicated if the defenders have a mobile force too, to keep the invaders occupied a bit. But even then eventually a window to attack the ground station will open.

Frankly, putting the planetary defense laser on a naval ship makes a lot more sense to me- at least it could move a little bit. And you'd still have the ocean as a heat sink. Or even on an aircraft, as the US Air Force does, but then one has to ask "then why not put it on a spaceship?"

2. Hidebound, Incurious, and Slave Mentality AIs as soldiers.

Again, I'll go with an extrapolation of actual reality. ALPHA had the better of every human pilot set against it. Every one. In ridiculously high ratios. So obviously, useable algorithms can be programmed into TL8 processors. Why on Earth not at TL10? Using sensors to find the enemy will be included in these algorithms.

I would say that Hidebound, Incurious, and Slave Mentality do not mean that one has to stand behind such a bot and give specific orders for every dependent task of a larger task. "Swing the pick at that rock. Now swing the pick at that rock again. Now swing the pick at that rock again. Now remove the rock fragments. Now brace the tunnel. Oh, sorry- now pick up that beam. Now place the beam vertically against the left wall. Now pick up that other beam." Etc. That's just absurd. One has to assume that a bot has some minimal programming meant to cover a task for which is it designed. Told to "dig a mine drift" a bot will not sieze up trying to decide which rock to swing the pick at first, either. At some point reason needs to get involved, here.

So, Hidebound, Incurious, and Slave Mentality do not mean that the bot marches mindlessly forward without even using its sensors to detect targets. What it might mean is that if told to "take that hill", it does so in a reasonable way, including scanning for threats beyond the confines of "the hill" that might impact its mission. It will not just march toward the hill ignoring enemies that shoot at it. It's a warbot, for heavens sake. But frex if it sees a cruise missile launch signature from the next valley over that in no way immediately impacts its current mission it might not determine that the missile launcher is nonetheless a very high-value target and decide to attack it itself, since it happens to be in a position that affords that opportunity, all on it's own initiative. It certainly would be programmed to pass the info up the chain, though. (I'm only just saying that to make the GURPS disads mean something, though. In reality, the algorithms would cover such things.) In short, such bots might not give themselves orders. But it does not mean that they are automatically incompetent at carrying out what orders they are given. And there most certainly will be something in their algorithm that covers new unexpected threats that pop up.

The Incurious part means that the bots do not read the trade journals and change their programming to cover new developments. They need a programmer to do that. Of course it also means that they don't pick up booby-trapped stuffed animals, either, so that's a plus.

The Hidebound part means that they don't develop new tactics- they just follow "the book"- but there is a "book". A bot might not think to drain a gasoline tanker into a bunker complex it has been told to clear, as the US did to Fort Drum in Manila harbor, if such wasn't in its programming. It would just perform its standard "assault on a hardened site" algorithm. Hidebound means that you can assume that "Ok, if there is a bot here, then there are probably others at 500m intervals to the left and right (within a built-in randomization factor) because that's how they are programmed". Because that's "the book" and they don't deviate from it unless ordered. It also might mean that if an ally betrays them and changes sides that they might struggle with identifying them as "enemy" until told to do so. Things like that. But even that might not be a problem, since any algorithm would have to cover enemies that are wearing friendly uniforms (or dressed as noncombatants) as a ruse de guerre.

That being said, "take over the world" is probably too broad for such a bot (i.e. an NAI). That might not go well.

[Rupert]

Originally Posted by Andreas

+9 isn't enough (and it should probably allow you to fire at a certain level of skill rather than give a bonus like that). It is very common for AIs in fiction to be very good at some of the tasks that computers are known to excel at. This is of course as you noted, a major world-building challenge, and misstakes are made, but it is possible to do reasonably well if you at least check that your future technology isn't worse than what could be developed today.

It is a conceit of GURPS that character skill matters. GURPS also assumes that professionals tend to have skill-11-12 in their main professional skills. Thus if something has been designed with an eye to balance at all, it will be designed to give the 'right' results (whether they be realistic results, or the most interesting results) with a skill-12 character making the skill checks.

If this offends your sensibilities, by all means simply assign effective skill levels to automatic weapon aiming systems, or just state that they never miss. I hope you want a game where the PCs are high level decision-makers in combat, because that what this will get you. It's also, in my pessimistic opinion, not terribly realistic even in open space, for things to never miss.

[Rupert]

Originally Posted by Michael Thayne

By "non-AI" do you mean "not statted as characters"? This is true of Traveller's gunnery programs. (Though the choice of terminology is a bit odd, since in the real world "AI" often refers to things that don't make much sense to stat as characters in GURPS.)

'AI' has had its meaning drift over the years, especially to computer programmers. Once upon a time it meant what GURPS uses it as, and what the general public would expect - a computer (or program) that is intelligent like a person or a higher animal. These days in the computing world it means all sorts of things, depending on context.

[Rupert]

Originally Posted by acrosome

I'll go with a continuation of modern reality as being most likely in the future- with force balance being anywhere nearly equal a nonmobile fortress will always fall to a mobile force. Think Eben Emael or the Maginot Line.

You're defending a planet. The invaders need to take it. Not sure in what way the Maginot Line is relevant. Eben Emael might be if you're suggesting that the invader will always find some novel way of assaulting, which I doubt.

The ships can alter their orbits to come over the horizon at an unpredictable time from multiple directions and then turn on the lasers to bathe the ground station, yet the ground station cannot do the converse.

But the ground based laser cannot use information gathered from a wide network of passive sensors across the whole planet? And the ground lasers, being bigger can't light up the invading ship from further out?

Then, the ships launch penetrators. If the ground station then opens it's shutters to PD the penetrators then it's at least a firepower kill, because the laser dies, and the penetrators hit anyway. And if it doesn't then the penetrators have at it unopposed, probably as a catastrophic kill. And it's already a mobility kill just by being a ground station. Consider that stopping a tank from being able to move, in modern combat, is considered a "kill."

Penetrators aren't free to carry, and they aren't free to deorbit.

Now, if the attacker is attacking with sufficient force, they can overwhelm the defence. If they don't, well they won't be landing forces today. Of course, if they have cheap delta-vee and just want to wreck the place, it's easier.

[David Johnston2]

Originally Posted by Andreas

Are there even any non-AI gunnery programs in those sources? You don't need any AI for aiming a laser in space. That is pretty much just a simple geometry problem! Rather than looking for AI skill levels, a better approach might be to just assume that there are easily availiablle non-AI programs with near perfect aiming capabilities, just like there would be numerous other programs which aren't mentioned in the books.

Oh sure. There is absolutely no reason for people to engage in violence whatsoever in any role other than supervisor or victim. Just let the mindless machines do all of it.

Except that sucks as an opportunity for gaming.

[acrosome]

Originally Posted by Rupert

But the ground based laser cannot use information gathered from a wide network of passive sensors across the whole planet? And the ground lasers, being bigger can't light up the invading ship from further out?
Penetrators aren't free to carry, and they aren't free to deorbit.

And sensors can't get degraded? It's a war, brother. They can't flood the stratosphere with hot chaff (or whatever)?

Originally Posted by Rupert

Penetrators aren't free to carry, and they aren't free to deorbit.

Deorbit doesn't mean "stop them dead directly over the target and let them fall." It just means "lower the orbit to impact the surface." That's much cheaper. Apollo did it using very little fuel.

Originally Posted by Rupert

You're defending a planet. The invaders need to take it. Not sure in what way the Maginot Line is relevant.

Ok. You're not sure. :)

Look, we can keep citing measures and countermeasures all day, and it's ENDLESS. But by the very nature of the nonmobile fortress there is one thing that you cede: mobility. You can't countermeasure that.

Originally Posted by David Johnston2

Oh sure. There is absolutely no reason for people to engage in violence whatsoever in any role other than supervisor or victim. Just let the mindless machines do all of it.

Except that sucks as an opportunity for gaming.

This at least is true. The Zeroth Law, and all.

[Ulzgoroth]

Originally Posted by acrosome

1. Laser ground stations vs. a mobile force.

I'll go with a continuation of modern reality as being most likely in the future- with force balance being anywhere nearly equal a nonmobile fortress will always fall to a mobile force. Think Eben Emael or the Maginot Line.

Or maybe don't, because you're grossly misrepresenting those. Eben Emael was defeated by a surprise commando attack from a direction (above) that its designers and defenders were inadequately prepared for. That gambit was critical to the german attack, because the fortress was fully expected to be a huge obstacle to their advance otherwise. Meanwhile, the Maginot line was surrendered mostly intact by political decision, after France had largely fallen behind it.

Would you care to tender an example that actually matches your principle? Or for that matter some argument to support it?

Originally Posted by acrosome

That orbital defense laser ground station that is being proposed as the most bestest thing evar is going to be meat. For one thing mirrors work both ways, so the orbital ships can just bathe it in laser light and then it can't open its presumably armored shutters, or the laser dies. The ships can alter their orbits to come over the horizon at an unpredictable time from multiple directions and then turn on the lasers to bathe the ground station, yet the ground station cannot do the converse. Then, the ships launch impactors. If the ground station then opens it's shutters to PD the impactors then it's at least a firepower kill, because the laser dies (and then the impactors hit anyway). And if it doesn't then the impactors attack unopposed, probably as a catastrophic kill. And it's already a mobility kill just by being a ground station. That's actually a decent point- consider that stopping a tank from being able to move, in modern combat, is considered a type of "kill."

Laser weapons probably aren't continuous, though some continuous beams sufficient to that sort of siege work might be worth carrying anyway.

No, you cannot come over the horizon at an unpredictable time, unless you've somehow cut off the laser emplacement from communication with the rest of the surface. Fixed emplacements would be connected to buried communications lines by which they would receive spotting information from observers all over the planet.

You could come over the horizon in large numbers at any one emplacement, but that's not particularly impressive since you were above the horizon to many other emplacements the entire time.

Also, none of this works against the proposed submarines.

Originally Posted by acrosome

For that matter, so long as your spies are even minimally competent, the people with the spaceships are going to know where the ground station is. Forever. So, they can arrange a swarm of relativistic impactors to hit it from all the way across the system (assuming that's possible in the setting, but even nonrelativistic but nonetheless really fast impactors will work) and the station can't dodge! Stealth them and put ECM on them (to the extent that you can, in space) and launch enough that the laser has no chance at PDing them all before they hit. Then, you don't even need the orbiting ships. They just come along later and drop the occupation troops.

That much at least is true: the problem that if your enemy controls the rest of your system and has unlimited time and delta-V to spend, they can conduct arbitrarily large kinetic bombardments at leisure from outside beam weapon range.

[AlexanderHowl]

The problem with relativistic weapons is that they are quite obvious to passive sensors during acceleration. Even a projectile going 0.1c will have a predictable trajectory and, unless their entire lift infrastructure has been eliminated, the defenders can get something in the way long before it gets to the target. A one gram piece of debris is the equivalent of 125 tons of TNT when a 0.1c projectile hits it, so anything will turn the projectile into a fireball. It would be a good use of packing peanuts...

[Ulzgoroth]

Originally Posted by AlexanderHowl

The problem with relativistic weapons is that they are quite obvious to passive sensors during acceleration. Even a projectile going 0.1c will have a predictable trajectory and, unless their entire lift infrastructure has been eliminated, the defenders can get something in the way long before it gets to the target. A one gram piece of debris is the equivalent of 125 tons of TNT when a 0.1c projectile hits it, so anything will turn the projectile into a fireball. It would be a good use of packing peanuts...

Shattering a .1c projectile isn't necessarily all that effective, unless you do it quite far out.

But that seems a bit afield.

[Rupert]

If the attackers are using 0.1c plus impactors, it suggests they're not interested in invading, and we're now talking about the best way to trash a planet at TL10, a different conversation.