I think GP's point is more relevant than your question implies. The vast majority of civilian flights could be flown entirely autonomously right now. (I'm not close enough to the aviation industry to make a confident guess, but +90% wouldn't surprise me.) Humans are there to take (decision or control) over when something goes off the happy path - in fact, pilots are encouraged / required to hand-fly landings that could be done automatically in order to keep their skills current. Obviously no one would accept even a 0.01% crash rate for civilian flights, so we're many orders of magnitude of improvement away from replacing pilots in that sector.
Military calculations are very different. Every military asset - most definitely including humans - is disposable, and all wars are (in some dimension) wars of attrition. Holding mission success constant, when the cost x capability x ability to manufacture for autonomous platforms becomes cheaper than all that plus training / replacement cost of human pilots, then human pilots will disappear. The logic of war being what it is, I expect HITL decision-making to very quickly be dropped as soon as it is seen to be retarding the progress of a cheaper option.
That's true. It'd be more accurate to say that the planes are capable, but that a lot about how the air transport system is organized would have to change to take full advantage of that - which we don't want/need to do, because an "automatic" system is too brittle, so we'd want skilled pilots on board, anyway.
I think GP's point is more relevant than your question implies. The vast majority of civilian flights could be flown entirely autonomously right now. (I'm not close enough to the aviation industry to make a confident guess, but +90% wouldn't surprise me.) Humans are there to take (decision or control) over when something goes off the happy path - in fact, pilots are encouraged / required to hand-fly landings that could be done automatically in order to keep their skills current. Obviously no one would accept even a 0.01% crash rate for civilian flights, so we're many orders of magnitude of improvement away from replacing pilots in that sector.
Military calculations are very different. Every military asset - most definitely including humans - is disposable, and all wars are (in some dimension) wars of attrition. Holding mission success constant, when the cost x capability x ability to manufacture for autonomous platforms becomes cheaper than all that plus training / replacement cost of human pilots, then human pilots will disappear. The logic of war being what it is, I expect HITL decision-making to very quickly be dropped as soon as it is seen to be retarding the progress of a cheaper option.
> The vast majority of civilian flights could be flown entirely autonomously right now.
This is extremely dependent on your definition of "autonomously" because at the moment they are not at all approaching autonomy.
At best, they'd be automatic, in the most rote way.
That's true. It'd be more accurate to say that the planes are capable, but that a lot about how the air transport system is organized would have to change to take full advantage of that - which we don't want/need to do, because an "automatic" system is too brittle, so we'd want skilled pilots on board, anyway.
That is why I was talking about civil aircraft. With the waiting list for those aircraft Airbus can stay relevant for a very long time.