A very interesting chat about the rapid development of military AI…
|
|||||
|
We are developing the social individualist meta-context for the future. From the very serious to the extremely frivolous... lets see what is on the mind of the Samizdata people. Samizdata, derived from Samizdat /n. - a system of clandestine publication of banned literature in the USSR [Russ.,= self-publishing house] Authors
Arts, Tech & CultureCivil LibertiesCommentary
EconomicsSamizdatistas |
How AI makes dogfighting drones unbeatableJuly 29th, 2023 |
 14 comments to How AI makes dogfighting drones unbeatable |
Who Are We?The Samizdata people are a bunch of sinister and heavily armed globalist illuminati who seek to infect the entire world with the values of personal liberty and several property. Amongst our many crimes is a sense of humour and the intermittent use of British spelling. We are also a varied group made up of social individualists, classical liberals, whigs, libertarians, extropians, futurists, ‘Porcupines’, Karl Popper fetishists, recovering neo-conservatives, crazed Ayn Rand worshipers, over-caffeinated Virginia Postrel devotees, witty Frédéric Bastiat wannabes, cypherpunks, minarchists, kritarchists and wild-eyed anarcho-capitalists from Britain, North America, Australia and Europe. CategoriesArchivesFeed This PageLink Icons   |
|||
All content on this website (including text, photographs, audio files, and any other original works), unless otherwise noted, is licensed under a Creative Commons License. |
|||||
The Zipper-Suited Sun Gods that are fighter jocks know, and have said in hushed-tones for years, that the last fighter pilots have already been born.
Yes and no. The ‘loyal wingman’ concept means there will still be fighter pilots, albeit rather less of them, but if they end up in a dogfight something has gone seriously wrong 😉
Yeah, tanks are obsolete, too…
All this really means is that there’s yet another notch on the ratchet between attack and defense that’s been clicked. Soon, you’ll have counter-measures, and then counter-counter-measures, followed by… Well, you get the idea.
Every time some “defense analyst” pronounces on something, about all you can count on is that they’re a.) entirely unaware of the history, and b.) dumb as f*ck.
College-educated, highly credentialed “defense analysts” told me, prior to 2003, that the US Army would never fight a war where we needed mine-protected vehicles in our rear areas, and that procuring armored route clearance equipment was a waste of money…
Anyone familiar with the history of these things could recount thousands of instances where “the experts” got it wrong. About all you can count on is that what they say, likely ain’t happening. And, what few times it does? It plays out in entirely unexpected ways. Also, after the fact? They’ll all stand around, patting each other on the back, telling the world what a good job they did in anticipating everything they failed to predict.
I’d be more concerned about the effectiveness of drone swarms. How could a single or pair of fighters deal effectively when up against a 1,000 drones which will do their utmost to crash into the fighters and blow up taking a wing, tail off. Perhaps we’ll see low power EMP weapons that can kill a swarm but then eventually we’ll get EMP-proof drones.
A F35 costs roughly $100 million. How many AI controlled drones can you get for that? Factory produced say at $50K each gives you 2,000 which is enough for several swarms. Swarms that can devastate an airfield, act as effective air defence or CAP.
Then there’s this. https://abc7news.com/killer-drone-viral-video-drones-simulation-facebook/2664856/
No ‘perhaps’ about it. Sustained ‘batch kill’ systems are already here in pre-mass-production form.
Not cheap ones you won’t.
Since you mentioned F-35s I thought I’d share something I just read:
America’s new F-35 fighter jets, for example, contain a magnet component made with an alloy almost exclusively manufactured in China.
What could go wrong?
A F35 costs roughly $100 million. How many AI controlled drones can you get for that? Factory produced say at $50K each gives you 2,000 which is enough for several swarms. Swarms that can devastate an airfield, act as effective air defence or CAP.
$50k is a different class of aircraft, and I doubt it is sufficient for combat (the satellite comms equipment probably costs more than 50k). The specific “loyal wingman” they mention, the Kratos XQ-58 Valkyrie, costs about $2M in large lots, or so the manufacturer claims. This is a MASSIVE cost reduction — basically an air wing for the price of an aircraft, but it ain’t $50k.
I’m not saying there won’t be things flying all over the battlefield. I’m just saying piloted fighters just won’t make sense once we get the last few bugs in the piloted and autonomous drone flees worked out. The fighter mafia won’t like it, and don’t like it now, but it is way cheaper in both dollars and lives to build disposable aircraft and train a pilot who can do his work in a cubicle instead of a cockpit. There will still be people involved in the whole process, they just won’t be pulling G’s and feeling the need, the need for speed.
That’s what they mean when they say the last fighter pilots have already been born.
@Steven R
to build disposable aircraft and train a pilot who can do his work in a cubicle instead of a cockpit.
That’s an entirely different thing that what the video was talking about. Remote control drones are already in wide use. And, FWIW, the large ones are not at all disposable, they are extremely expensive. The risk with them though is communications disruption, what happens if the man in the cubicle can’t fly the plane because he can’t talk to it? The reason for human pilots is that they are autonomous systems who can make their own decisions in isolation even without communication. It is the development of autonomous AI systems that challenge that. There is no “pilot in a cubicle” for them, just a pre-programmed mission, and the ability to execute that mission without further communication. Though one configuration as mentioned, is a commander flying one plane and several drones under his control.
I know what they mean but they’re still not quite right. There will be “manned fighter” missions within a loyal wingman context, particularly as the EW environment becomes less permissive & over the horizon coms get harder to use. Manned gun/off-the-rail engagement will likely be rare but I’d be surprised if the fighter-as-command-hub will not be armed for just that eventuality because shit happens.
However, all-drone pre-programmed attacks is just another way of saying “long range cruise missile strike”, just with AI special sauce, not a new thing at all. The difference between a “drone” & a Storm Shadow is fairly arbitrary.
That’s all broadly correct, Fraser.
@Guy Gibson’s Cancelled Dog
BTW, I like you user name. He was a great man and a it was great movie. Though I’m glad you didn’t use the dog’s actual name… we’d have all been in trouble then.
See Star Trek Beyond for the effectiveness of high-quantity small-size attack automata.
Defence saturation isn’t a new concept, but swarm = cheap. Cheap = vulnerability to electromagnetic effects or just rapid ROF cannon fire or lasers or whatever. Cheap also = (relatively) slow. Hardening against electromagnetic effects/lasers/whatever = not cheap any more. Fast (as in ‘fast jet fast’) also costs money.
Drone swarm is a legitimate approach & it will get some traction, but it’s not a new paradigm or a wunderwaffe. Swarm attacks will probably be an issue more at the lower tiers of conflict rather than peer/near peer high intensity first world participant wars. Also has potential as a poor man’s first strike approach before your enemy is set up fully. Hezbollah have kind of been trying this out against Israel & Israel has responded in typical first world style, by throwing tech, money & smarts at the problem with some success as they are well aware that Iron Dome causes the problem to change shape rather than ends it.