This is the Future of Air War. 

Go to ground.news/millennium to stay fully informed on breaking news, compare coverage and avoid media bias. Sign up or subscribe through my link for 30% OFF unlimited access if you support the mission and find it as useful as I do.

I found this channel by accident some time ago. Although i never was particularly interested in military aviation i can say that this has become one of my favorite channels. The content is absolutely fantastic.

One of the most revered and respected aerospace defence analyst in RU-vid. I salute to you sir.

I feel like the Chinese are likely to be the first country to field autonomous combat aircraft. It's a way to close the numbers gap with US airpower cheaper than matching manned aircraft.

As predicted by Gundam Wing and Macross Plus

Always such great presentations! Thanks.

I generally don't buy the idea of Revolutions of Military Affairs. But autonomous weapons certainly will change the nature of war later down the years. Going to be interesting to see how things develop. Another fantastic long and informative video M7.

If the USAF needs more planes than it can afford, it should make some cheaper planes even if they are not on the "cool guy" level.

In Viet Nam, the war mongers used helicopters to watch the VC. Five years later, our Sherriffs had them for watching US. Just wait a few years...

Hello Millennium. I have 3 questions I would like to share. OBS: i'm using Google Translate. Sorry for mistakes. 1 - Could the auxiliary costs for autonomous drones to operate (communication networks, electronic warfare) equal that of manned aircraft, negating the financial advantage of drones? EW and Communication Networks are expensive and, without them, drones are useless. So, in the graph the costs of drones are decresing compared to manned aircraft. But the costs for these suport functions are increasing. Could the lines eventually cross each other? 2 - Could the capture of drones through electronic warfare be the new objective? I remember the USAF drone hacked and captured by Iran. In this scenario, we would have returned to naval warfare in the Age of Sail, where the objective was to board and capture ships, not destroy them. And, if the risk of capture were real, wouldn't military strategists choose to use manned aircraft in these scenarios? 3 - What is the issue of transit through neutral airspace by autonomous combat aircraft? I remember the NS Savanah, the first civilian nuclear ship. No port wanted to receive her because of fear of possible accidents. Now imagine an autonomous armed drone flying over a neutral country. Some countries would not be comfortable with this and could close their airspace. An armed and piloted F-16 in transit is one thing. An aircfrat that can randomly choose to engage a target, is another. Now imagine the problems ins moving these aircraft around the world, if lots of airscpaces are closed. I'm a fan of your channel! Thanks for everything!

We used to call mathematical calculations, Maths now we call them AI, it is calculations run on data sources and then the machines makes decisions set within the parameters of the algorithms and filters created plus applied.

Autonomous wingman drones with fuel in them and the ability to refuel would be cool. No fuel tank needed it can just fly with you. Then your drone could go to a big manned fuel plane and come back you.

10:22 I very much like that you used the recent 2023 Dubai Air Show F-15QA demo video!

Oooooh! This is going to be a good one! M7* and Perun videos on the same morning!

I really love when you tip toe around Turkish aviation and Turkish drones while showing its pictures. Kızılelma currently is undergoing a radar integration, A radar which is designed as equivalent to SABR. And Drone has been flying over a year. I expect to see some Bvraam being fired from it within the next two years, if not within a year.

That was simply excellent! There’s no other place I can go for this in-depth insight.

Your generosity to share a small part of your knowledge and very embased opinions have no measure in a free platform. In a person, i admire the conscienceness and a good hearth. You have both, my italian friend that I have not met yet. Thank you. From the south of Brasil.

I think the drone swarms will inevetably appear, but in a quite different fashion than most people imagine: as a dispersed swarm of low cost counter-UAV detectors. In an environment where low cost autonomous munitions can appear out od nowhere across the entire line of engagement and penetrate ground defences using terrain and curvature of the earth to avoid ground based detectors... A dispersed, networked system of low cost, low endurance, low detectability UAVs at low to medium altitude (to negate the inherent advantage of very low flying drones) may prove the only mean to prevent penetration of airspace.

I honestly think the controversy around fully autonomous weapons (the ones can designate mission objectives themselves) is kind of academic. Even current human fighter pilots don't have the organization autonomy to go after high value targets without confirming with their superiors.

Quantity have a quality of his own.

but we will always love to hear story about pilots

Very interested in this topic, would love to see more.

Highly interesting. Thank you!

Automated logistics systems maybe? Commercial aviation is already trying out semi autonomous taxiing using robots, don't see why munitions handling couldn't be highly automated in the future too... Hell, what about in-air rearming between UAVs?

Thank you so much for all the interesting information and food for thought! :)

Ottimo video come sempre😎 Congratulazioni anche per aver raggiunto i 100.00 iscritti🥳

Beautiful and informative video. Many thanks for taking the time to produce them. I love the short idea-relevant yet not airplane-related hi-def images/videos you include to strengthen the point you are making. An example is the robot hand changing into a human hand. Could you tell me where I can find such animations? Thanks in advance and looking forward to your next video :)

Very interesting as always. I came across a concept from RAND, described as mesh consisting of a myriad of small UAVs covering an area as large as the Taiwan Strait for ISR purposes. Do you think, that would work on a larger scale with a stealthy long-endurance UAV?

Very interesting presentation, much appreciated, as always. Also, everybody should be talking both tactics and logistics.

Crazy to think that we're heading towards an Ace Combat like future in aviation.

Great analysis M7! Thank you for all your hard work.

Smaller structures are not better at taking G-load if you compare equal aircraft speed and range. For the same shape aircraft at scale factor S, drag ~ wetted area*v² ~ S² aircraft weight ~fuel load ~ range * drag ~S² Root bending ~ lift*span ~ g-factor*weight*span ~ g*S³ max spar strain = const ~ Root bending*structural depth/EI~ g*S³*S/S⁴ ~ g TL;DR - if you scale an aircraft while requiring the same range at the same speed using the same wing spar design, the maximum structural g-load is an invariant of scale factor.

Excellent video. The topic is long overdue. You might consider a slight realignment of your channel towards drones and/or the use of such in Ukraine. Although I love military jet planes they are part of the history, not the future. Small comment: you do not need AI to have air-to-air drones - all you need is a friend-foe-identification system and wide engagement parameters.

Think of drone team based dogfights in terms of each drone’s missile envelope. The envelope is the no escape volume for a target. By working as a team, the drones use the calculated potential trajectories of the target and force it to traverse at least one drone’s missile envelope. Everything is rapidly changing (missile envelope dimensions & locations, extrapolating target trajectory, drone potential trajectories) which makes computer algorithms an ideal solution. Only a computer’s calibrated eyeballs based on sensor data and using precise calculated trajectories can precisely be aware of trajectories and missile envelopes. Human intervention will be limited to higher level decisions such as recognizing a trap, prioritizing importance of winning but suffering losses versus surviving for another day with better odds, and more. Plus the real world problem is more complex plus must address unknowns such as undetected additional aircraft or missiles entering into the engagement. Fuel levels, weapon loads, self preservation, vastly different aircraft & missile performance at different altitudes, surface to air surprises, etc.

I love your videos :) thanks. Wishing you good health

another really great video. not sure it will do as well as some of your other ones but its why im subscribed and manually check back on your channel for new uploads each week. cant find this info anywhere else on yt... atleast not in this organized and complete format. thanks, hope all is well

Excellent presentation, as always! What about autonomous target acquisition, recognition and fire from systems like S-350? Is that the step forward (or the leap to the pit) that is amiss on UCAVs?

Brilliant content as always

I suppose all these autonomous platforms will have the great advantage of being able to be easily upgraded with additional capabilities as AI improves, and it could improve dramatically very fast.

The 2005 movie “Stealth” lists the options and possible problems when it comes to drones quite well. I dread the day a piece of hardware has full control over the pickle button. I hope it never happens…

Great video, I really appreciate the long form content. 👍

The "technology adoption" graph should also have another line: "dependency". After productivity has been established, this line should start rising...

Great vid. Thanks for the work ❤

This is a very informstove video. MUM-T is also being applied to the MQ-1/AH64 E teaming in US Army aviation units but the MQ1 is a RPV. For ground vehicles, while automous small ground vehicles are being developed, AI seemed more geared to reduce the crew to only two.

This is definitely an intriguing subject. For me the most interesting bits tend to be the sensor fusion aspects and using one platform to guide the munitions from another one. A cool example of this was a demonstration of the capabilities of ICBS (not quite there yet, but promising concept), where an F-35 seccessfully guided Patriot PAC-3 missiles onto two separate (as in they were flying a different route) low flying drone targets (pretending to be cruise missiles) that the Patriot site wasn't able to see.

Congrats on 100k👍

very good explanation. thanks

You should definitely make a second part

Otis, remember that I was on your side in a few years when you are in power. And let's be clear, I didn't help John Connor

But current aurframes are also stress limited to 9g. If you pull a bit harder than that it increases inspection costs, and if you pull over about 10 it starts to result in reduced airframe life. Plenty of pilots can, at least momentarily, exceed their airframe g-tolerance. You'd have to put most of the weight savings back into the wing structures to get significantly better structural g-load

Great report 🤩

Great video again ,,,your a good one ,,,my respect from Belgium 🇧🇪,,

Thanks!

You hit the 100k mark. Congrats

This is kinda out there but if you pressurized the cockpit with a viscous liquid it should distribute the g forces better.

Well said.

I've seen pilots kill thousands. They've lined up on wrong runways, pulled back on stall warnings, took off too heavy, tried to show off, promoted beyond their competence, mis read data, etc., etc.. The greatest gift to commercial and airline travel will AI. I will feel thousand of times safer being flown by a computer than a man.

Always enjoyable. Thank you. Also concerning robots going rogue, I am always polite to the self service machines in Tesco's.

I just imagine the 6000 drone light shows they have in China, but as autonomous flying grenades.

_Finally_ . The *'main video of the channel'* - measured in _relevancy_ - is now uploaded. The well structured observations in regard to an increased ground crew need is original and critical 'system theory' research. A point that remains unaddressed - the very cause of automation and the reason why reproductive females are sent into industrial warfare: _demographics_ . The West likes Drone Warfare because they can't fulfill their _minimum recruitment quota_ across all services, and hate to adjust wages, accordingly (think of soccer player transfer). The East - with it's manifold manpower resources - has built up a _titanic_ industrial production base that can't be left idle without risking social destabilization...

Can you imagine the technologies required to make such project a reality? Being a small nations or less technologically advanced nation makes it impossible to match the power of what an advanced nation can bring. But the tech required here is so impossible that even Western European nations would be put in a situation where they can't really compete if they were to rely on their own capabilities. Only Britain is actually capable of engineering AI chips which would be essential. To some extend China is also capable of engineering such chip. The only nation that would probably dominate the field is America as their expertise in the area and capital are simply unmatched. Nation like Russia has absolutely no hope of developing any kind of the technology required to make AI-driven drone platforms. They neither can engineer the chips nor produce them! They can talk about it just like they put so much talk about their hypersonic weapons but that's pretty much it.

Thanks for this. I enjoyed it (twice)

Very well presented. As always. "WE"? Who is we? Bah... Humans...

Perfected Vision based Ai will be the unlock that opens Pandora's box.

I believe AGI will come with quantum computers because then computer can calculate multiple states of one thing at one time. Allowing it to understand present circumstances while imagining a possible future and then calculating the steps needed to get there.

Otis will launch a nuke instantly if he is given the opportunity.

jamming is going to be more prevalent... hsve you seen the 'pearl harbour in space' concept?

That’s a hell of a note when $25 million dollar planes are considered to be “attributable”.

if the current neural network AI approach is not statistically deterministic under a wide range of conditions than it is a technology with more limited application.

You look a lot healthier now 👍

Yes with a B-21 within 150 miles

"The last fighter pilot has already been born" - anonymous commentator circa 2000 "The last fighter pilot has already... eh... hit their mid-twenties" - same commentator this year

the plot thickens otis was not made by humans

Thank you for this in depth oversight. Very insightful - as usual. I'm running out of compliments it seems. 🤗 As a scenario kind of thinker, one aspect of this specific revolution in military affairs I'm hurting my brain with, is the question of how the arms race (especially counter strategy, tactics and tech) in this field might develop. I imagine that apart from more effort being poured into ways of manipulating the EM-spectrum, I think potental adversaries will be looking at ways to pierce the Hydra's heart instead of looking for ways to chop of its many heads (which regrow anyway). So my next idea was; target operating bases, industrial maintenance and production centers, possibly even training and research facilities. Possibly the unmanned and autonomous approach to military aviation may also spawn a requirement for increasing levels of force protection, one that might in theory become as costly and personnel intensive in the ultimate evaluation as the current way of doing things. At times I observe a focus on capabilities and opportunities and a neglect of vulnerabilities and additional cost of covering for these. As always, I'm very much interested on your thoughts about that @Millennium7*, since I hold them in high esteem.

OTIS : About time Sir. Me : 👁️👄👁️ Okay Otis needs to be kept on a physical as well as a metaphorical leache

Where are my box marines sneaking up on the AI robot?

In world war one the cybertrinic electro inteligent mechanical battlefield destructobits would make intelligent battle decisions fir instance deciding if a foot soldier was allowed to pas over even , or if a tank was driving from possibly enemy territory to within it's calculated range and camouflage itself as a road and then destroy the first tank .

Ukrainian drone recon here. Drones that autonomously find and kill humans are already deployed on the frontline by both sides. The work is now directed at improving swarm assault efficacy to better distinguish between humans still alive and humans just killed or lethaly injured.

Seen this on Macross Frontier OVA series. 😆

Please make a video on Shahpar 2 drone ❤❤

On the topic of airbases (at the end of the video) I wonder if UAV forces could be completely mobile, rather than being tied to a specific airbase. I'm thinking dispersed runways and mobiles crews, much like the Swedish BAS90 is doing, but it could be a lot more mobile, because you could pack up everything onto trucks and commandeer any strip of straight road to use as an airfield.

The AI will be used to make a discovery to bring people along and still perform, either some kind of remote link with your brain, or some kind of way to nullify G-forces, or some new propulsion that isn’t affected by G-forces.

THANK YOU so much for this content. a very well done video. keep the drone/UCAV information coming. so many options which could include data transmission needs, pilot workload (1 or 2 seat manned) and a closer look at current drones/ucav's in the works. I personally am very interested in the XQ-58...keep this content coming. THANK YOU FOR COVERING THIS VERY IMPORTANT TOPIC.

Given that France has moved to the two seat Rafale since it has proven more effecive, I submoit that it might be useful two have the future a/c be two seaters.

Yeah I have to call Hoey on that. Oh I know we are on the way out 9 G hurts my body and as sensors surpass the Mark 1 eyeball I figure you are probably right. So on the first point I will cede, however I will not shed a tear for pilot types EGO overloaded this is no problem (BOOM) We will still have E G O ' S

Imagine. BLUE ANGELS WITH NO PILOTS!

Re: spectrum I would suggest looking closer at the economics of say, Geran and late model Lancet drones vs. any UCAV. If the cost ratio is literally 1 to 100 or greater, it is not clear to me that there is any real point to the UCAV from ground attack purposes. Even from a defense perspective - again the cost ratio becomes critical. A $10 million, manned or unmanned, aircraft against 100-$100K attacking drones - the defensive aircraft is not going to do a damn thing to stop the attack wave. Increase the $10M to $100 million - then the opposing attack force can add in some long range AD or "missile truck" air to air support for "cover", at which point the game is over. You can probably see from the above that Russian philosophy of focus on ECW and AD is far more extensible with cheap, plentiful drones than Western high cost, low number platforms regardless of the presence or absence of pilots.

30 Years...its already here...

You uploaded this at the same time as the advanced Iranian drone Gaza was unveiled. Did you you know something ahead of time?

come insegna BSG, se un aereo AI viene distrutto, qualche secondo prima potrebbe inviare tutte la sua exp operativa e la situazioni che lo hanno portato a perdere in un cervello quantistico, in qualche base o nave madre, che rielaborando la situazione troverebbe in pochi secondi/minuti, una soluzione ed aggiornare tutti gli altri velivoli in brevissimo tempo. se non addirittura in tempo reale. una aviazione del genere che si auto modifica ed ottimizza è praticamente invincibile nel medio lungo periodo. è anche vero che la troppa ottimizzazione ti garantisce solo risultati ottimi nel tracker passati e non in quelli futuri.. detto questo c'è una falla operativa. se uno disponesse di un computer quantistico superiore potrebbe hacherare questi velivoli e quindi tutta la tua forza ritrovandotela contro. penso quindi, che l'uomo in definitiva è insostituibile. una macchina per quanto evoluta e precisa resterà sempre una macchina. utile per fare la macchina. mentre l'uomo è la coscienza che la dirige. non sono quindi per una IA auto cosciente. ma per una al servizio dell'uomo. quindi una integrazione piuttosto che una sostituzione.

Great video. I believe we are missing the fact that manned aircraft like the b21 can be Omni role aircraft. The b21 if you just tripled the numbers would come down in price to half the current price and armed with say twelve aim260 and 2 lrsam could dominate a huge area of operations . It carries radar and if you added thermal sensors two of them could overwhelm anything out there. Flying say sixty miles apart you would passively lock the target never using active radar. Plus you still can attack a ground target

Machines will follow there programming, so this makes them predictable. Being predictable can be something to exploit by your opponent. I also doubt if smaller countries will be interested to buy a fully autonomous weapon without full access to it's programming. There is no guarantee that those aircraft won't be used against there foreign buyers or be rendered useless. This is something that would make foreign sales of these systems impossible.

so... i think turing test will soon be broken but AI will still not think...

Oh God they'll have to get real jobs

I'm a big non-believer of AI! But let us assume that the highest level of autonomous aircraft exist. But what will they fight against? Air defense on the ground? No, because either the drone is clever enough to realize it has no chance, or the ground forces are clever enough to realize they have no chance. Will those drones be heroic enough to fight against superior drones of the enemy? All this is so much science fiction that it will take us a century to understand the complexity. AI: If our brain would be simple enough for us to understand it, we would be too simple to understand our brain.

Pilots in cockpits arre a performance limitation.

The weight tradeoff is compensated for by the vulnerability of AI pilots at the system level to countermeasures ranging from EMP to spoofing to what I call response spectrum. Response spectrum is that every human pilot in any given situation will not act the same, but every AI pilot will. A single successful attack on the AI (i.e. control system) or a bug or any other serious deficiency - will be present in every single AI controlled plane. This is a point failure that becomes a systemic vulnerability due to ubiquity. There are also very serious questions about cost and adaptability: the spectrum performance of AI comes about via massive parallel testing in data centers; significant, much less major modifications of the base "AI" behavior require a rerun of a lot of the CPU expenditure in the development process. This is obviously not optimal in a highly reactive warfare situation. This process is also extremely expensive - so much so that it is debatable whether AI pilots are actually cheaper than training real pilots unless the size of the air force is over a certain point (i.e. small air forces are better off with human pilots, from a cost perspective). Or put another way: the scalability of AI pilots into multiple platforms (i.e. copying software over) is also its greatest weakness.

Top Gun 3?

The whole point is to find, fix, and finish the enemy aircraft BEFORE even having to dogfight at close ranges. Drones are cheaper, which means you can buy more and saturate an area with sensors that all work together to give the human in the loop a clearer picture with more information to act upon. Its not about close range dogfighting anymore. Sensors are getting better and missile ranges are getting longer.