The solution to pilot shortage?

This forum has been developed to discuss aviation related topics.

Moderators: sky's the limit, sepia, Sulako, lilfssister, North Shore, I WAS Birddog

av8ts
Rank 8
Rank 8
Posts: 848
Joined: Fri Apr 08, 2011 8:31 am

Re: The solution to pilot shortage?

Post by av8ts »

The bots will still be on here
---------- ADS -----------
 
User avatar
complexintentions
Rank 10
Rank 10
Posts: 2183
Joined: Thu Aug 19, 2004 3:49 pm
Location: of my pants is unknown.

Re: The solution to pilot shortage?

Post by complexintentions »

This again. *sigh*

I agree with Panama Jack. I find the attempts to extrapolate past advances to future ones, absurd. I also think Rockie nails it a few posts earlier. To believe that autonomous aircraft are just around the corner is to display ignorance as to the sheer amount of human judgement required on each and every flight.

Analogue DVD’s to digital streaming is not a valid analogy to pilotless aircraft. Horse-with-rider to car-with-driver, is not analogous to piloted-to-pilotless aircraft. The automation of land-based vehicles, operating in two dimensions with the luxury of braking to a halt, is not remotely the same thing as aircraft, traveling in three dimensions and unable to stop mid-flight (other than in very unfortunate ways). Three pilots to two pilots is not the same as two to one, or removing them altogether, when considering redundancy. "They don't have to be perfect, just better than humans." Not a very high bar to clear with most human car drivers! A little different when examining aviation accident rates versus auto accident rates, or comparing the loss of life in a single aircraft accident versus a single automobile one. The point is these articles are riddled with false analogies and inflated premises. Sheer number of clickbait links does not make them more “true”. A robot that makes (with human help) more precise sutures than a surgeon is proof the surgeon is nearly obsolete? Give me a break.

Who designs, builds, and programs the machines, computers and software that take over from the humans? That's right - other humans. Shifting the weight of "human error" from pilots to software engineers is a lateral move (that's being charitable), not a move forward. And even if one day, how to create something smarter than the creator is somehow resolved, ceding control to such machines poses far bigger issues than worrying about losing pilot jobs. But put me in the camp of being skeptical that imperfect humans can create perfection in machine form. It’s antithetical to both the laws of logic and entropy. Human error is human error, with software it just gets embedded in the code. Silicon Valley hubris aside.

I'm not foolish enough to discount the pace of technological advance. But I've yet to see anything remotely compelling in any of these breathless articles predicting the imminent wide-scale arrival of completely autonomous passenger aircraft. Just saying, if they want to make their point, they'll have to do far better than holding up driverless trains and Google cars and robot arms that throw switches as "proof" of autonomous airliners. None of these operate without human decision-making and oversight SOMEWHERE in the process nor at anything resembling an acceptable level of reliability without it. The article quoted makes vague references to "recent advancements in aeronautics". Since when is the obstacle to autonomous flight a question of aeronautics? The field of artificial intelligence is far more relevant. The aeronautics have been largely solved decades ago. It speaks to the credibility of this popular meme.

Personally I think that rapid technological advances in energy sources, aerodynamics, engines, avionics, materials, man-machine interface, or a whole new yet-unforeseen form of transportation that renders aviation, not just pilots, obsolete, are all far more likely that an as-yet not technically possible, prohibitively expensive conversion of the entire present system for theoretical improvements in safety and highly questionable economics. Hell, the A380 will NEVER turn a profit, and it's one machine, current generation! Hysterical reactions to things like AC 759 (how many people died there, again?) tend to distort the fact that aviation is already an almost absurdly safe way to travel. Entrusting it entirely to software will make it even safer? Does anyone who believes that actually USE computers? The return on investment just isn't there, not for safety gains, not for economics.

In other words...I don't see pilotless, autonomous aircraft any time soon as wholesale replacement to the current system. Sure, niches like military drones already exist, and will continue to expand. And I am not saying "never" to the fanciful scenarios envisioned in some of these articles. But first it has to be technologically possible. Which would require true AI - not sophisticated algorithms and sensors, but true self-aware machines. Not there yet. Maybe we'll get there, maybe we won’t, it's certainly not a foregone conclusion. But then it would also have to be economically advantageous to the point of redesigning the entire infrastructure. An even bigger obstacle. All this must happen in a timeframe that precludes being leapfrogged entirely by any new transportation technology - the pace of technological advance also applies to competing technologies, after all. And lastly it presumes that the whole world is preoccupied with similar goals as ours, when the truth is vast swaths of the planet are still focused on trying to find enough food to survive another day.

Meh. I'm still waiting for the flying cars in every driveway we were promised in the 50's. Completely technically feasible, so where’s mine? :mrgreen:
---------- ADS -----------
 
I’m still waiting for my white male privilege membership card. Must have gotten lost in the mail.
GATRKGA
Rank 4
Rank 4
Posts: 217
Joined: Fri Aug 04, 2017 1:24 pm

Re: The solution to pilot shortage?

Post by GATRKGA »

complexintentions wrote:This again. *sigh*

I agree with Panama Jack. I find the attempts to extrapolate past advances to future ones, absurd. I also think Rockie nails it a few posts earlier. To believe that autonomous aircraft are just around the corner is to display ignorance as to the sheer amount of human judgement required on each and every flight.

Analogue DVD’s to digital streaming is not a valid analogy to pilotless aircraft. Horse-with-rider to car-with-driver, is not analogous to piloted-to-pilotless aircraft. The automation of land-based vehicles, operating in two dimensions with the luxury of braking to a halt, is not remotely the same thing as aircraft, traveling in three dimensions and unable to stop mid-flight (other than in very unfortunate ways). Three pilots to two pilots is not the same as two to one, or removing them altogether, when considering redundancy. "They don't have to be perfect, just better than humans." Not a very high bar to clear with most human car drivers! A little different when examining aviation accident rates versus auto accident rates, or comparing the loss of life in a single aircraft accident versus a single automobile one. The point is these articles are riddled with false analogies and inflated premises. Sheer number of clickbait links does not make them more “true”. A robot that makes (with human help) more precise sutures than a surgeon is proof the surgeon is nearly obsolete? Give me a break.

Who designs, builds, and programs the machines, computers and software that take over from the humans? That's right - other humans. Shifting the weight of "human error" from pilots to software engineers is a lateral move (that's being charitable), not a move forward. And even if one day, how to create something smarter than the creator is somehow resolved, ceding control to such machines poses far bigger issues than worrying about losing pilot jobs. But put me in the camp of being skeptical that imperfect humans can create perfection in machine form. It’s antithetical to both the laws of logic and entropy. Human error is human error, with software it just gets embedded in the code. Silicon Valley hubris aside.

I'm not foolish enough to discount the pace of technological advance. But I've yet to see anything remotely compelling in any of these breathless articles predicting the imminent wide-scale arrival of completely autonomous passenger aircraft. Just saying, if they want to make their point, they'll have to do far better than holding up driverless trains and Google cars and robot arms that throw switches as "proof" of autonomous airliners. None of these operate without human decision-making and oversight SOMEWHERE in the process nor at anything resembling an acceptable level of reliability without it. The article quoted makes vague references to "recent advancements in aeronautics". Since when is the obstacle to autonomous flight a question of aeronautics? The field of artificial intelligence is far more relevant. The aeronautics have been largely solved decades ago. It speaks to the credibility of this popular meme.

Personally I think that rapid technological advances in energy sources, aerodynamics, engines, avionics, materials, man-machine interface, or a whole new yet-unforeseen form of transportation that renders aviation, not just pilots, obsolete, are all far more likely that an as-yet not technically possible, prohibitively expensive conversion of the entire present system for theoretical improvements in safety and highly questionable economics. Hell, the A380 will NEVER turn a profit, and it's one machine, current generation! Hysterical reactions to things like AC 759 (how many people died there, again?) tend to distort the fact that aviation is already an almost absurdly safe way to travel. Entrusting it entirely to software will make it even safer? Does anyone who believes that actually USE computers? The return on investment just isn't there, not for safety gains, not for economics.

In other words...I don't see pilotless, autonomous aircraft any time soon as wholesale replacement to the current system. Sure, niches like military drones already exist, and will continue to expand. And I am not saying "never" to the fanciful scenarios envisioned in some of these articles. But first it has to be technologically possible. Which would require true AI - not sophisticated algorithms and sensors, but true self-aware machines. Not there yet. Maybe we'll get there, maybe we won’t, it's certainly not a foregone conclusion. But then it would also have to be economically advantageous to the point of redesigning the entire infrastructure. An even bigger obstacle. All this must happen in a timeframe that precludes being leapfrogged entirely by any new transportation technology - the pace of technological advance also applies to competing technologies, after all. And lastly it presumes that the whole world is preoccupied with similar goals as ours, when the truth is vast swaths of the planet are still focused on trying to find enough food to survive another day.

Meh. I'm still waiting for the flying cars in every driveway we were promised in the 50's. Completely technically feasible, so where’s mine? :mrgreen:
That was one phenomenal post. :shock: / slow clap.
---------- ADS -----------
 
av8ts
Rank 8
Rank 8
Posts: 848
Joined: Fri Apr 08, 2011 8:31 am

Re: The solution to pilot shortage?

Post by av8ts »

+1
---------- ADS -----------
 
rxl
Rank 7
Rank 7
Posts: 691
Joined: Tue Jan 08, 2013 6:17 am
Location: Terminal 4

Re: The solution to pilot shortage?

Post by rxl »

complexintentions wrote: But first it has to be technologically possible. Which would require true AI - not sophisticated algorithms and sensors, but true self-aware machines. Not there yet. Maybe we'll get there, maybe we won’t, it's certainly not a foregone conclusion.
Great post Complex! I chose to quote these sentences of yours because it gets to the heart of what worries me the most about the "brave new world" that the AI types are racing all of us headlong towards. Like you said - "Maybe we'll get there, maybe we won't" - but at some tipping point "It" WILL in all likelihood get there and at that point do we really know what the outcome will be? Up to that point, we have been in control and we have been the "creators". After that, all human bets are off. Do we really want self-aware, self-motivated machines?
I think we need to think long and hard about the long term outcome of this.
---------- ADS -----------
 
Post Reply

Return to “General Comments”