One step closer
One step closer to Skynet.
It's only one plane now, we still have a chance!
The US Air Force Test Pilot School and the Defense Advanced Research Projects Agency (DARPA) claim to have achieved a breakthrough in machine learning by demonstrating that AI software can fly a modified F-16 fighter jet in a dogfight against human pilots. The claims rest on the USAF and DARPA implementing machine learning in …
Indeed.
Plenty dogfighting since WW2 also...the Sea Harrier was doing a mighty fine job of that thing during the Falklands conflict in 1982 (actually a bit scary when you realise how long ago 1982 actually was, but certainly a lot less than the century that the OP refers to)
Dogfighting is not a supersonic activity. Aircraft travelling at supersonic speeds tend not to change direction. Supersonic performance is for getting to and getting out of the combat zone.
Harriers used their trump card “VIFF” very effectively on the rare occasions they were involved in combat.
werdsmith,
If I remember correctly, VIFFing is something that's cool for air shows - but not something the RAF or Fleet Air Arm ever used in combat - or even had in their doctrine. The problem with putting the brakes on in a dogfight is that you probably can't get that speed back again, and I'd imagine the Harrier isn't the fastest accelerating plane. But a small use of the thrust vectoring might give you a big advantage in turning.
However the OP's post above is still stupid, and reductive. The reason Harrier did so well is a combination of factors. Almost all the Harrier's kills were shots from behind, which their older Sidewinders could already do, because it was doctrine to shoot from behind even with the better missile. Also, I'm not even sure that there were enough of the Sidewinder L's to go around anyway, so not all kills were achieved with them, and not everyone had access to them.
But also the Argentinian planes were at the edge of their range, and didn't have fuel or time for dogfighting. Plus were mostly on bombing missions anyway. They had to fly lower because of the threat of Sea Dart - which wasn't as good as advertised - but still dangerous to anything at medium and high level. Also the British were getting raid warning from the submarines and Chile, as well as the air defence picket ships.
Finally after the Black Buck raid on the Islands' runway - that killed any idea of forward-deploying fast jets. But also the Argentinians got worried that the RAF might try the same on the air bases in Argentina itself - which meant they withdrew some of the Mirage fighters for defence. Which made the Harrier's job a bit easier.
I don't believe that to be the case.
As soon as you remove the human from the risk then the appetite for those in command to make less will thought out choices increases.
Fighting from the safety of a concrete box is essentially a war of attrition in terms of who runs out of assets first if both sides have the same tech. Once you have finished that you are into conventional warfare however the issue is that if you have committed too much to the first stage you are royally f****d. This is the most likely outcome as manufacturers will lobby for what makes them the most money.
The more pressing concern is that the use of drones, remote controlled or relying on AI results in decisions being made that are far more likely to result in escalation if things kick of badly.
Equally we have seen many scenarios recently where all the military resources and tech that is available has been defeated by low-tech groups driving around in a Hilux armed with a 50 cal, RPG, Stinger or a few blocks of C4.
Most of those groups will be able to access some form of fissionable, chemical or bio material and make a mess very easily in a large city if they feel they have nothing to lose.
Fighting from the safety of a concrete box is essentially a war of attrition in terms of who runs out of assets first if both sides have the same tech.
Basically the 'cannon fodder' model as seen in WW1, except that the fodder can easily be replaced by the military-industrial complex (thus acting as a nice little earner for the manufacturers)
No matter how much and how sophisticated hardware you can throw into a war, you'll still lose to the side that's prepared to throw in all that plus humans.
Drones are part of the arsenal, but they haven't replaced humans. Nor will they, unless we one day reach a stage where nothing a human can do will ever influence the outcome.
"No matter how much and how sophisticated hardware you can throw into a war, you'll still lose to the side that's prepared to throw in all that plus humans."
Not true. And Poland how things went for them when Hitler invaded. Poland fielded a WW1 army with cannons and horses versus modern German tanks and artillery.
Poland couldn't match the German hardware, so the comparison doesn't apply anyway.
It's only in the past decade that drone technology has reached the point where some commanders are starting to dream of bloodless (on their side at least) wars. So far, it hasn't gone great.
"Humans plus drones" beats "just drones". Ask the Ukrainians.
>Not true.
It IS true.
Eg Russia vs Germany, late WWII. Quite literally exhausted German defence by soaking it up with dead men. Then simply marching over them.
Eg Otto Skorzeny tells the story of a horror episode where the Russians just kept marching into a sighted-in artillery chokepoint. Each unit utterly marmalised. He'd been commanding the defence, after a few days noticed the guns got patchy, drove up to find the CO (leutnant) nissed as a pewt and having a mental breakdown. Outrage! Sent him to the lockup and took over the unit himself. To discover for himself the horror.
As he put it, by the end of the day, each fresh unit was "wading through strawberry jam" up to their knees. The jam was their predecessors. Whom they'd watched march into the chokepoint. And joined them.
That night, Skorzeny released the leutnant, apologised, turned the unit back over to him.
But the Germans soon ran out of shells, and had to retreat.
Russia in WWII actually managed to exceed the most extreme stupidities of WWI. By a long way. And routinely.
Swarming is a valid tactic. And extremely effective. All you need is to disregard the humans as people
And Poland how things went for them when Hitler invaded. Poland fielded a WW1 army with cannons and horses versus modern German tanks and artillery.
Greg38,
Germany also fielded a mostly WWI army against Poland. With infantry, horse-drawn artillery and mule trains for supply. The Germans invaded with 60 divisions of which I think 7 were Panzer divisions and 6 were motorised. The rest walked. Germany also had at least one cavalry brigade. They did have better aircraft though.
The only fully mechanised army in the world at the time was the British.
Incidentally I recently heard an old BBC interview with Ian Fleming - where he said that the reason he left officer training at Sandhurst was that he saw this mechanisation in the 30s as turning him and his fellow officers into mere motor mechanics. Actually he was thrown out for conduct unbecoming an officer having contracted gonorrhea. Who said James Bond wasn't based on real life...?
However it doesn't matter how good your air force is, or how shiny your drones - if you want to take and hold ground, you need the poor bloody infantry to do it. Everything else is just supporting them in that role. Similarly at sea, if you want sea-control you've got to have ships. Plus the supporting arms to keep those ships alive. You can deny the sea to the enemy, with submarines, drones and/or aircraft. But if you want to do useful things, like move supply or invade with troops - you need to be able to keep ships there.
Ghengis Khan, whose armies conquered most of Asia only regretted attacking the Sami, who just retreated North. Khan's armies did not know how to survive in the frozen cold wastes (and the Sami did not two how to fight a 'regular' army). Khan's armies basically died of cold or retreated. There was never a 'front line' just a trail of reindeer in the snow leading their enemies to oblivion. I'm sure Sun Tzu would have appreciated their strategy.
An_Old_Dog,
Upvote for Hellburner. Great book. All the "Company War" books are good - although some are more depressing than others. Looking at you 40,000 in Gehenna.
I suspect though, that in dogfighting there are a lot of moves that are simply the best counter to another - given the aerodynamic performance of the two jets in question. And so this is something that may be done by computers as well as people.
However the problem might be that making a superior jet, that can pull 20g and still be useable the next day, might end up being as expensive as making a 6th generation stealth jet and crewing it with 2. At which point these high-end AI drones are going to be anything but expendable. Or your loyal wingmen are basically going to be indistinguishable from missiles.
"why would anyone want to get up close when you have dozens of stand-off weapons at your disposal"
Maybe ask some Vietnam veterans about that... just because you have weapons in your arsenal doesn't always mean you'll be permitted to use them.
Also, as low observable technology continues to spread, and the risk increases of engaging in combat with an adversary your BVR weapons struggle to lock onto, you may find that regardless of what the rules of engagement say about using such weaponry, you might still not have a choice in the matter.
This reminds me a little bit of when a US Air Force AI drone 'killed operator, attacked comms towers in simulation', but then hadn't, as it turned out.
Your comment prompted a recollection of this fun tale, too
Instead of retrofitting a meatbags vehicle, task AI to build its own. Missing a trick there, I'll await my cheque in the post.
But seriously, pick 100 random people and tell me which one is a "bad guy". Then ask someone else and chances are, you'll get a different answer.
Let's stop here and stop using this tech to line the pockets of the few and instead, further mankind as a whole.
If we aren't going to do that, just press the big red button already cos I don't like this timeline.
Agreed - in my view one of those major issues is that it's just a few short mental leaps for the military to consider that if the AI can dogfight, couldn't it use the aircraft's missiles as well? Then the Powers That Be could have humans out-of-the-loop in a combat air patrol.
And then you get this again: Iran Air Flight 655 [wikipedia]
Ultimately, we'll have autonomous fighters on both sides -- expensive machines destroying expensive machines. A 1966 episode of Star Trek (A taste of Armageddon) presented a more cost effective solution -- simulated battles fought entirely on computers. The only snag was that the notional casualties on both sides were obliged to report to disintegrators as soon as the battles were over.
Unfortunately, for all the automation, war still primarily consists of human attrition one way or another until one side gives up.
Star Trek also gives us a few examples of where AI warfare may lead
Prototype (Star Trek: Voyager) (When the two planets called a truce and attempted to terminate the robots, the robots destroyed their creators out of self-preservation and continued their war)
The Arsenal of Freedom (Star Trek: The Next Generation) - When a weapon sales AI doesn't take "no thanks" for an answer
"Star Trek also gives us a few examples of where AI warfare may lead"
However there remain vast bumbers of purely human ways to initiate and wage war without the need for "AI". It's a sad fact that some war or other has been waged somewhere on this planet continuously for a very long time, mostly using old fashioned technologies such as sharp things and bullets.
veti,
Why build expensive machines to destroy expensive machines, when a score of $500 drones can do the same job for a zillionth of the cost?
Because they can't. Look what happened when Iran lobbed 300 mixed drones, ballistic and cruise missiles at Israel. Almost all of them were shot down by expensive fighters and expensive air defence systems.
I guess you could argue that the SAMs shooting down drones and missiles is already computer-on-computer. But there's no AI hype to be squeezed from that.
Also the $500 drones are incredibly low capability and quite easy to shoot down. The Shahed's that Iran, Russia and the Houthis regularly lob around are more like $70,000 a go. And they're also pretty limited. The question is, can you overwhelm better systems by having lots of them?
This is why the Royal Navy are now rushing their anti-missile laser (Dragonfire) into front line deployment - as well as lots of countries looking at ground based laser or microwave weapons and maybe the return of ground based autocannons or multi-barreled guns for the anti-drone role.
With all the warning the defenders had (the drones and missiles came from Iran, some distance away from Israel), they didn't manage to shoot all of them down.
Had Iran really wanted to escalate and not just make a token gesture, they would have been able to overwhelm the Israeli defense.
Note also that the cost of an anti-missile missile is far above the one for a drone or even a basic missile.
Had Iran really wanted…
Probably the largest single barrage of ballistic missiles ever. Maybe Iran could overwhelm Israel’s defences. Maybe not. Expensive though. Obviously the drones are cheap. But long range cruise missiles cost more than SAMs. You don’t use the expensive long range ones to shoot them down, because you can’t see them at long range. Ballistic missiles ain’t cheap either.
It was a major escalation, by the way. Changing from a proxy war, to a massive direct act of war.
If I understand well a computer has beaten a human in a computer simulation. All we need to do now is to convince the $BADGUYS to fight against us in a simulator, instead of cutting-off our supply-chains, or any other unpleasant real-world events, and our victory is certain. Well done DARPA !
There is no info about any cameras or sensors for the AI to see the other fighters, only 3D images on a simulator screen. This looks like an AI trained on real flight data but the "dogfight" seems to happen in a computer simulation.
I believe Killer Robots are inevitable and will be produced in the coming years. The argument will be that they can "Save Lives" or enable smaller countries to pack a bigger punch.
Mind you that Killer Robots will also enable genocide since the Germans introduced the gas chambers in WWII as a "humane" measure. Not for the victims, but for their soldiers, who sustained psychic trauma from all the killings and executions. With Killer Robots you could simply say: "Kill everyone in this city, men, women, children, everyone!" and the Killer Robots would comply without mercy.
How many dog fights have Israel and Iran had these last few days ?
How many dog fights has there been over Ukraine and Russia ?
ZERO.
We are in a new dreadnaught age, where suddenly over night all the OLD ships are worthless, the original Dreadnaught was the start of the end of the British Empire, because suddenly its superiority became nothing because old ships were worthless.
Missiles and drones is where its at, you cant shoot down 20 missiles or 200 drones with an F16 or F35.
Yes, this sort of thing was mentioned in an episode of The 3 Body Problem, that the ship commander guy wasn't too enthusiastic about the ship (especially its fuel consumption) and he said if he was in charge he'd not gave made a fancy expensive ship but would have commissioned a load of drones instead.
Makes sense, though. Ships are big, slow, intimidating, and need lots of people to make it work. But it's about fuck all use of your enemy has a couple of hundred self piloting drones with explosives attached. We've seen (in new years light show displays) the sort of control a drone can manage. A bunch of those things weaponised could be extremely damaging if inflicted against an unsuspecting populace. Set it to home in by GPS, fly below anybody's radars, and deploy at the same time, and it'll all be over before anybody knows what "it" even is.
Warships? Fighter jets? What use will they be? Those are things from an era when the enemy was countries. Half the time these days the enemy is an ideology with no geographic boundaries.
werdsmith,
Anti-swarm defences have been being researched for ages. there are Already quite a lot in place. The Royal Navy's latest Type 31 doesn't even have a big main gun. It's got a 76mm and 2 x 40mm cannon for the purpose of being able to deal with boat and aerial drone swarms - as well as the higher end SeaCeptor missiles for medium range air defence. The RN have multiple sets of Phalanx CIWS (multi-barrel close-in defence) as well, which they fit to ships according to the threat level - which I assume can also go on T31. They've also announced they're going to start fitting Dragonfire lasers to the fleet as well.
We are in a new dreadnaught age, where suddenly over night all the OLD ships are worthless
No we're not. HMS Dreadnough made everything older obsolete, because you can't easily change the armour or turret scheme on an early 20th century warship. They'd sometimes remove a turret, to fit other stuff, or fit up-rated guns - and in World War II there was always room on deck for one more anti-aircraft cannon. But the fundamentals of the ship were done when you built it. HMS Hood had a bit of deck armour added for example, but there was never the time or budget for a full re-design, and even then you could change the deck armour or add things like torpedo bulges - you couldn't make a modern King George V class out of a WWI fast battleship.
But look at something like the Royal Navy's type 45. It was built to have room for expansion. Which is the RN's new design philosophy. All warships must have physical room for more stuff, plus spare power generation capacity for new weapons and electronic warfare.
They're a 20 year-old design. So the ships are undergoing a mid-life upgrade to give them better engines with more power. And sadly also to correct a problem with the original design of the powerplant - that had them breaking down too often. That's PIP.
There's also the ballistic missile defence upgrade to the SeaViper (Aster/PAAMS) surface-to-air missiles. France and Italy have already done this for the land version of SAMP/D, which France have given one set of to Ukraine. So it already had limited ABM defences, it's shot down Iranian Houthi ballistic missiles in the Red Sea. But the missiles are also being upgraded for much faster missiles. Thihs may or may not require further upgrades to the radar - but there are already radar upgrades of some kind in the works.
They're also replacing the short range Aster/SeaViper missiles with SeaCeptror (CAMM) - but doing it by adding another 24 VLS cells. So it's going from 48 (all now long-range) to 72 SAMs. They're also adding another 8 cell launcher for NSM (Naval Strike Missile) - which is a mid-range norwegian cruise/anti-ship missile. There was also the option to make that a 24 cell Mark 41 VLS launcher, compatible with all the US weapons - but I think that went out on cost/availabilty grounds. But could be done if they really feel the need, it just means removing the gym - the space allocated for this purpose. If we did that, it would mean you could have quad-packed SeaCeptor and so have another 48 of them - and still have 12 cells free for Tomahawk or the Storm Shadow replacement due to start arriving in the next year or two.
Plus it's got the power for the Dragonfire laser and/or microwave anti-drone weapons. Plus other electronic warfare capabilities.
You could even rip out most of the SAMs and fill it up with anti-ship or land strike cruise missiles and turn it into a completely different class of ship. And if it turns out that your current SAM of choice is crap - you can just update the systems to use another.
This kind of flexibility means that you can easily operate a ship like this for 40 years, and it can be cutting edge in technology terms for its whole lifetime. As long as you keep your eye on the ball and keep the upgrades coming. It's probably the best air defence ship in the world, just because it's so tall. Meaning its radar horizon is higher, so it's much better at dealing with sea skimming missiles than any other AAW ship - even though it doesn't have the best ABM capacity (though improving quickly).
Spartacus: No we're not. HMS Dreadnough made everything older obsolete, because you can't easily change the armour or turret scheme on an early 20th century warship.
cow: Exactly. Figher planes are obsolete just in a different way. 100 drones or 100 missiles against an F15 or F35 will result in the same things happening, 90 odd will get thru - its that simple.
Spartacus: Plus it's got the power for the Dragonfire laser and/or microwave anti-drone weapons
cow: Dragonfire cannot shoot anything down, it simply warms up the target in the sky, and you hope it fails and falls from the sky. Secondly all lasers need time to charge and cool down between uses. WHen the DragonFire is positioned on the Ukraine border shooting down 50 Iranian drones then you can pretend its perfect today .
CowHorseFrog,
You have asserted that fighter planes are obsolete with literally zero evidence.
As events at the weekend just proved, fighters can easily intercept slow-moving and cheap drones. This is one of the reasons we've been trying (horribly slowly) to get F16s to Ukraine. They don't have the kit, the mass or the training to take on the Russian air force and ground based air defence - but they may be able to gain tactical air superiority locally for short periods of time in order to be able to support their ground forces. And in the meantime they can shoot down lots of Shaheds - and protect Ukraine's cities and power infrastructure.
But the things that a NATO strike package can do with electronic warfare and a few hundred fighter aircraft backed by tankers and intelligence and reconaissance aircraft are massively more effective than what you can do with a few hundred drones.
It's possible that drone technology will make most aircraft obsolete. But we're not even close yet.
ICBMs are of course extremely hard to stop (I'm not sure I'd say impossible). But they're also incredibly expensive, as are other hypersonic missiles. The great thing about a plane is that you can use it more than once. The other problem with replacing aircraft with ICBMs is that you might know you've put a conventional warhead on the end of it, but the target might not. So you might find a nuclear response being returned before your conventional warhead has even arrived.
Finally, the fact that laser and microwave weapons are now being deployed, not just tested, suggests to me that they may be at least somewhat effective. We'll find out. But for now, they'll be part of a layered defence, which is how warships operate anyway. Most major warships have SAMs, often of different ranges, plus longer range guns and then short range self-defence cannon. The RN are now adding lasers, the US Navy are currently deploying anti-drone microwave weapons to their ships for drone defence.
There are also a huge bunch of short range, cheaper missiles that are either being looked at to re-purpose for anti-drone work or in design. So the US are looking at designing a half-sized air-to-air missile, for short range anti-drone engagements. So aircraft can carry twice as many. The UK have been testing Martlet and Starstreak in this role, Starstreak is a man-portable SAM already being used in Ukraine in a vehicle mounted anti-drone role. Martlet is based on the same system, but is for non-armoured vehicles or bunkers. However that's also been used in Ukraine as a cheap weapon to shoot down drones and the RN already use it on helicopters in the anti-drone boat role. There's been testing done for a while now to make a naval version, that could be used against drone boat and aircraft swarms.
Most new weapons can be countered. The secret to success in warfare is good training and having the right mix of different weapons systems available to use in combination to get what you want done.
spart: As events at the weekend just proved, fighters can easily intercept slow-moving and cheap drones. This is one of the reasons we've been trying (horribly slowly) to get F16s to Ukraine. T
cow: The F16 are not dog fighting in Ukraine, they are launching the UK + FR air missiles.
Feel free to show me a news story about a dog fight of any kind in Ukraine, because there are zero, and if you do find one, i wouldnt call that a success either because most o fthe damage on both sides is done by missiles and drones overwhelming the other side.
spart: Most new weapons can be countered. The secret to success in warfare is good training and having the right mix of different weapons systems available to use in combination to get what you want done.
cow: Lots of words, answering questions i never asked and in no way addressing my original statement.
try and stay on topic.
spartacus: You have asserted that fighter planes are obsolete with literally zero evidence
Cow: I give you the past 50 years.
Basically the last real dog fights happened in the Vietnam war.
One dog fight in Libya or Iraq shows how little effect they have on those two examples.
spartacus: But the things that a NATO strike package can do with electronic warfare and a few hundred fighter aircraft backed by tankers and intelligence and reconaissance aircraft are massively more effective than what you can do with a few hundred drones.
cow: Firstly fighter planes have a shitty range. THeres no way America is going to send hundreds of fighters over Russia because they dont have the range. THey cant even fly from Nato bases to Moscow and back.
Fighters only work against enemies with basically no or joke air forces like iraq or afghanistan.
Notice America hasnt got the balls to fly over Russia or China and ha snot done so for a long time.
spartacus: But look at something like the Royal Navy's type 45. It was built to have room for expansion. Which is the RN's new design philosophy. All warships must have physical room for more stuff, plus spare power generation capacity for new weapons and electronic warfare.
Cow: Of course they plan... that doesnt mean im going to pretend everything is perfect.
Facts are planes in the sky have been obsolete for a long time, they only work against donkey riders in places without an airforce against anyone else they are worthless. Intercontintental ballistic missiles have been unstoppable for 50 years
Spartacus:
No we're not. HMS Dreadnough made everything older obsolete, because you can't easily change the armour or turret scheme on an early 20th century warship. They'd sometimes remove a turret, to fit other stuff, or fit up-rated guns - and in World War II there was always room on deck for one more anti-aircraft cannon. But the fundamentals of the ship were done when you built it. HMS Hood had a bit of deck armour added for example, but there was never the time or budget for a full re-design, and even then you could change the deck armour or add things like torpedo bulges - you couldn't make a modern King George V class out of a WWI fast battleship.
Cow:
Modern missile systems have made planes obsolete for different reasons.
A F35 is never going to stop 50 missiles.
Except for one or two dog fights which happened by chance like the TOmcats over Libya, nobody has tried to actually have an air war with dog fights for over 50 years.
That bullsht is over.
If America ever has a war with China or Russia, dog fights arent going to be a deciding factor, because of nukes.
Spartacus: Plus it's got the power for the Dragonfire laser and/or microwave anti-drone weapons. Plus other electronic warfare capabilities.
cow: YOu just verified my statement.
The machines dont need pilots dogfighting its all missiles flying many times faster for a lot less money.
<<<<<JET FIGHTER>>>>>>
I am not the only person becoming increasingly fed up with the ignorance and poor education, coupled with a low standard of English which leads almost every media article you see nowadays to twist nomenclature round into a nonsensical order.
It is a Jet Fighter, not a fighter jet - the jet is merely the power source that drives it. You wouldn't refer to an airliner jet, or a car motor, or a bike push. Why twist it the wrong way round for a jet fighter? It's no different from the brain dead arseholes back in the 1970s who insisted on referring to the Reliant Robin as a robin-reliant.
As soon as I see it referred to as a 'fighter jet', I immediately know that I am dealing with someone who hasn't got much idea what they are talking about!
I've heard loads of RAF and USAF pilots say, "when you're flying the jet" or whatever. They do call them jets. And although I've never heard airliner jet, I have heard passenger jet plenty of times.
I don't believe there's any rule of language here. Merely a choice of style. And style guides are not defnitive, since everyone's disagrees. Even dictionaries get updated for how people actually use the language.
Therefore I deem your complaint rediculous, and predict that you will loose this argument.*
*My apologies. I couldn't resist.
Couldn't spell either :) - the word is spelled 'ridiculous'; no e present.
There is a rule, or at least a whole set of them which determine in which order words should be used, and the fact that various people are illeducated and get the order wrong is not a valid justification for claiming that using the words in the wrong order is OK. The 'jet' is the engine, the source of the motive power, it is not the noun which refers to the vehicle itself, and therefore should be placed before the noun (which in this case is actually aeroplane, becuse we are talking about a jet powered fighter aeroplane, which has become shortened to simply 'fighter').
Changes to language are good where they make the language richer, clearer and more expressive. Too many people seem unable to appreciate the difference between evolution of a language and deterioration of a language, and the current frequency of the use of the term 'fighter jet' falls firmly into the latter category.
When a dedicated aircraft with no cockpit or human support systems, and airframe stressed to twenty or twenty-five G's is produced, no manned aircraft will defeat it.
But it'll cost even more than a manned jet, and probably have even more stringent maintenance requirements. Meaning nobody will be able to afford many of them.