It is hard to determine what is “useful” or “important”. In case of emergency or chaos, moving info around brings confusions.
> Air traffic controllers think on the fly more than you would assume. They look at their radar scopes and plan each target to work with others by working their speed, altitude, and path. Computers could probably work out these, but then how would they relay that information to pilots?
> Another thing to think about is the vast number of VFR aircraft who are not on a flight plan going A to B. ATC is able to communicate with them, learn their intentions "we saw something neat and would like to orbit here for 5 minutes", and help them see and avoid other aircraft.
> The fact that they are humans who you can talk to is tremendously helpful in emergencies as well. If I'm fighting a sick airplane I might not have the time or focus to find the nearest suitable airport, get weather information and important notices, and plan a descent. A controller watching over me sure can though.
Humans are flexible in a way computers can't be yet.
-----
> All aviation-related technology is tightly regulated and goes through an immense amount of testing and consideration before use.
> All of the obstacles we have around self-driving cars-- "whose fault is it if it crashes", "how do we know it's safe enough before letting it on the road", "do you still need a licensed driver", etc-- are 10x more problematic in the sky. It isn't necessarily harder to automate, but is harder to declare "officially safe enough".
It's not enough for an automated ATC to be somewhat good or decent. There is little margin for error when the consequences of such an error could be the death of hundreds of people and the loss of a multimillion dollar plane. ATC in real life is much more difficult than it appears to casual observers.
It just has to be as good as a human, and GPT-4 is already smarter than most humans.
ATC in particular depends on exact phrasing, in issuing commands, acknowledgement, and repetition back, so GPT’s ability to fudge something nice sounding is actually pretty bad.
In some pre-defined contexts.
Essentially all of that can be changed to exchanging information digitally and automatically. In fact most of that information exists digitally. It's not a technical challenge to do that. There are many ways to make this happen. The challenge is entirely institutional because it would represent massive changes to an industry that resists all form of change for decades on end. And that's for good reason: safety. But it also means that a lot of very flaky technology keeps on getting used and that pilots are task saturated with a ridiculous amount of work during safety critical portions of flight. When they get behind on their tasks, dangerous situations happen. When they stop practicing, they lose the ability to stay on top of things. It's that hard. And their pilot license shortly after (currency is not optional).
If you'd design the procedures and communication from scratch using modern tools, you'd end up with a very different system than what we have today. The fun thing indeed is that with modern AI, what might happen instead is that we end up with AI assisted ATCs and pilots delegating the ritualistic exchange of information via VHF radios to AIs. They could just send each other digital messages instead that are then processed automatically and acknowledged manually when needed. But on the off chance of one of the pilots or atcs still being stuck in the nineteen sixties, it's going to be easier to pretend that everyone is and have AIs imitate human actors participating in this.
So, text to speech and speech to text on both ends combined with an AI that has enough contextual information to take simple decisions and control systems escalating to pilots and controllers only when needed to might be more doable than coming up with a new system. I'd say chat gpt 4 would be well up for a "pretend you are an ATC controller" type role playing session.
CPDLC https://en.wikipedia.org/wiki/Controller%E2%80%93pilot_data_...
Isn't approach one of those times where 1) you don't want the pilot distracted by more visual information (instead of audio), and 2) pilots listen to exchanges with other pilots to keep situational awareness?
You can define rules how planes should and shouldn't fly, verifying that this is always possible would be a nightmare.
ATC getting overloaded with more traffic is a given either way. Loads of electrical drones will start showing up in the next decades. Some piloted, some autonomous. ATC will have to adjust to that. And keeping everything the same with more human controllers is not going to scale.
Entirely possible we could find more things to move there with humans