We'd known since the late 60s when
Firebee's were beating up on F-4's that manned aircraft would have operational issues against unmanned ones and AI was going to be another major milestone on that road. When you consider how far you tend to have to "dumb-down" AI's to give human game players a chance it's a sobering thought but on the other hand that's within a highly controlled and restricted environment where the AI has most of the advantages whereas in an actual environment with limited sensors...
The fun thing there is that if you look into the video's origins (I got bored once and watched the whole presentation before which was five hours long) you actually find out that the AI is being nerfed even there - it doesn't have something like a 360 degree view as you might think it does (ie, in the simulation it isn't just a floating ball of cameras), it is being fed similar information to Banger, the pilot for Team Man in this contest of Man V Machine. What it has is an advantage in that it is able to process all this information simultaneously, so whereas he has to do things like look at his instrument panel to know stuff like his air speed, the AI always has this kind of information in its virtual head. It basically has a sensory advantage over him as a result in that it can absorb all this information at the same time, which makes for a hefty advantage, but the real killer is to do with the precision of maneuver; both Huron and Banger had the same plane with the exact same load out and thus the same performance, but we consistently see Huron outmaneuver Banger over and over again, and the reason for that is to do with the "tightness" of its maneuvers. Because it has perfect information coming in and knows the exact state of the airframe and all performance metrics, its performance in a maneuver is basically 99% accurate, whereas Banger just can't keep up with it. To explain that perhaps a bit better, consider something like rally car races - Huron is making every gear change at the exact moment they need to make them to get the maximum possible performance out of the car, whilst Banger, performing at his best, simply can't do it so precisely.
The result is a few percent of an advantage in this simulation that consistently results in him getting his ass whooped, and even his last tactic of diving to the deck with super tight turns only drew out the fight rather than won it. I wrote about that video in the past, so I can just quote myself here...
The first is that this is not a real world situation, which would make the situation even worse for a Human pilot - that last set of maneuvers that drew the fight out as long as it did were, as the announcers said, constant 9g turns at 500~ mph, an action that the Human body just cannot sustain for long, whereas an AI can keep that up for as long as the mechanical components of the airplane can...something that can last far, far longer than the fleshy-bits of a Human being.
...and point out that Banger himself was also gaming the simulation by making use of how he'd not actually be affected by the simulator G-load. The real nails in the coffin, though, come from a Navy pilot, Commander Colin "Farva" Price, who wrote this. I've bolded the best bits:
It does not take much skill to put the aircraft’s lift-vector on the other aircraft and yank on the Gs. In fact, if in doubt, just doing that will take care of 75 percent of the fight. But BFM is about being smoothly aggressive. Understanding the difference between when it is necessary to max-perform the aircraft and when it is time to preserve or efficiently gain energy back is key. In a tight turning fight, gaining a couple of angles at each merge can suddenly result in one aircraft saddled in the other aircraft’s control zone working a comfortable rear quarter gun-tracking shot.
In true gamesmanship fashion, the guns-only BFM engagement was the setting for the AlphaDogfight contest. So what jumped out at me about the engagements? Three main points. First was the aggressive use of accurate forward quarter gun employment. Second, was the AI’s efficient use of energy. Lastly was the AI’s ability to maintain high-performance turns.
During BFM engagements, we use training rules to keep aircrew and aircraft safe. An example of this is using a hard deck, which is usually 5,000 feet above the ground. Aircraft can fight down to this pretend ground level and if an aircraft goes below the hard deck, they are considered a “rocks kill” and the fight is ended. The 5,000 feet of separation from the actual ground provides a safety margin during training.
Another training rule is forward-quarter gunshots are prohibited. There is a high potential for a mid-air collision if aircraft are pointing at each other trying to employ their guns. Due to the lack of ability to train to forward-quarter gunshots, it is not in most aviators combat habit patterns approaching the merge to employ such a tactic. Even so, it would be a low probability shot.
A pilot must simultaneously and continuously solve for plane-of-motion, range, and lead for a successful gun employment. It is difficult enough for a heart of the envelope rear-quarter tracking shot while also concentrating on controlling a low amount of closure and staying above the hard deck. At the high rates of closure normal for a neutral head-on merge, a gun envelope would be available for around three seconds. Three seconds of intense concentration to track, assess, and shoot, while at the same time avoiding hitting the other aircraft. The Heron Systems AI on several occasions was able to rapidly fine-tune a tracking solution and employ its simulated gun in this fashion. Additionally, AI would not waste any brain cells on self-preservation approaching the merge avoiding the other aircraft. It would just happen. The tracking, assessing, and employing process for a missile is not much different than the gun. I am pretty confident AI could shoot a valid missile shot faster than I can, given the same data I am currently presented within the cockpit.
The second advantage of AI was its ability to maintain an efficient energy state and lift vector placement. BFM flights certainly instill aviators with confidence in flying their aircraft aggressively in all regimes of the flight envelope. However, in today’s prevalent fly-by-wire aircraft, there is less aircraft feel providing feedback to the pilot. It takes a consistent instrument scan to check the aircraft is at the correct G, airspeed, or angle-of-attack for the given situation.
Even proficient aviators have to use a percentage of their concentration (i.e. situation awareness) on not over-performing or under-performing the aircraft. AI could easily track this task and would most likely never bleed airspeed or altitude excessively, preserving vital potential and kinetic energy while also fine-tuning lift vector placement on the other aircraft to continue the fight if required.
Lastly is AI’s freedom from human physiological limitations. During the last engagement, both aircraft were in a prolonged two-circle fight at 9 Gs on the deck. A two-circle fight is also referred to as a 'rate fight.' The winner is the aircraft who can track its nose faster around the circle, which is directly proportional (disregarding other tools such as thrust vectoring) to the amount of Gs being pulled. More Gs means a faster turn rate. 9 Gs is extremely taxing on the body, which the pilot in the contest did not have to deal with, either. A human pilot would have to squeeze every muscle in the legs and abdominals in addition to focused breathing in order to not blackout. During training, I maintained 9 Gs in the centrifuge for about 30 seconds. Then I went home and took a nap, and that was without being shot at. AI does not care about positive or negative Gs. It will perform the aircraft at the level required.
But the critical thing to remember, the most important thing to remember, is that Huron isn't the end of this line of development - it is just the end of the driveway, and the real destination is still very far away. Technology like that is only going to improve further and further, and just leave Human pilots behind in the dust because machines are just flat out better at raw calculations than Human beings are, and ultimately that's what aerial warfare is when you strip down all its niceties. How many degrees to get that gun shot, how much air speed will I lose on that turn, how many meters per second will my missile travel before it hits the target, so forth and so on. A computer is just better at that kind of stuff, and in an environment like a plane, that's a lethal advantage. The military in general agrees with me on this one, which is why we're starting to see stuff like drone wingmen starting to appear, like Boeing's own, which is a completely autonomous wingman designed to support a Human pilot, but the step from "support" to "replace" isn't that far. The Ghost Bat...
...is already rated to be ready for autonomous flight, so we're already heading to the tipping point in that regard, but we're getting off topic here by moving from tanks to air planes
2) Attached automation, aka remotes to repair and upgrade the main unit
I was going to mention that myself, but figured it best to leave that sort of thing as an exercise for the reader!