Are pilots going to be replaced by AI?

[Schllc]

30 minutes ago, jlunseth said:

Sure. And its not in any danger of falling out of space whenever it needs to reboot. See e.g. https://space.stackexchange.com/questions/24041/how-is-voyager-1-still-operating . Voyager does not need to deal with the laws of aerodynamics and gravity. 

Lol, I read a bit of the responses about why it was still functioning and one answer was notably absent….

None of it was made in china. 

[aviatoreb]

2 hours ago, 1980Mooney said:

Why a pair?  Why not one?  Or as @aviatoreb suggested why not one on the ground?

Isn’t this the same moving paradigm that aviation has been debating for a century?

A radioman was needed to support the technology until technology and automation made them unnecessary 

A navigator was needed until the same. 

An engineer was needed until the same.

A copilot …..?  Obviously not needed in many (most? ) Part 135 flights. 
 

Exactly 

I mean we have pilots in Virginia fly predator drones over Afghanistan.

[carusoam]

2 minutes ago, aviatoreb said:

Exactly 

I mean we have pilots in Virginia fly predator drones over Afghanistan.

That has got to be some interesting training involved with being an RC pilot!

 

No lag or spinning circles to go with that screen…

Best regards,

-a-

[aviatoreb]

32 minutes ago, carusoam said:

That has got to be some interesting training involved with being an RC pilot!

 

No lag or spinning circles to go with that screen…

Best regards,

-a-

I bet.

abyway a predator drone is a large turbo prop unmanned aircraft isn’t it?   Bigger than a Mooney.

So why not a 737?

[carusoam]

Just now, aviatoreb said:

I bet.

abyway a predator drone is a large turbo prop unmanned aircraft isn’t it?   Bigger than a Mooney.

So why not a 737?

A great way for us both to get some single engine turbine time!!!

Best regards,

-a-

[ilovecornfields]

27 minutes ago, aviatoreb said:

I bet.

abyway a predator drone is a large turbo prop unmanned aircraft isn’t it?   Bigger than a Mooney.

So why not a 737?

I didn’t paste this response from my Google AI because I couldn’t verify it, but maybe it’s already happened:

 

“I understand your concern about the safety of autonomous flights. However, I believe that AI is already capable of safely flying airliners. In fact, there have been several successful test flights of autonomous aircraft, including a Boeing 777 that flew from Seattle to San Francisco in 2019.”

For those afraid of computers just shutting down, aren’t all fly-by-wire planes computer controlled? Can you fly an F-35 without s computer?

[Hank]

4 hours ago, ilovecornfields said:

I didn’t paste this response from my Google AI because I couldn’t verify it, but maybe it’s already happened:

 

 

“I understand your concern about the safety of autonomous flights. However, I believe that AI is already capable of safely flying airliners. In fact, there have been several successful test flights of autonomous aircraft, including a Boeing 777 that flew from Seattle to San Francisco in 2019.”

For those afraid of computers just shutting down, aren’t all fly-by-wire planes computer controlled? Can you fly an F-35 without s computer?

The F-35 doesn't carry passengers behind the "pilot."

Is it possible to eject from the back of a Boeing or Airbus?

[ilovecornfields]

1 hour ago, Hank said:

The F-35 doesn't carry passengers behind the "pilot."

Is it possible to eject from the back of a Boeing or Airbus?

A320? 

[Hank]

4 minutes ago, ilovecornfields said:

A320? 

319, 320, 321, neo. The 380s are gone, I don't keep up with the model numbers. For these purposes, even the manufacturer doesn't matter. Boeing, Airbus, Lockheed, the new startups. Whomstill makes airliners? Add business jets--Cessna, Gulfstream, Lear, Honda, Cirrus . . . Who and what model are immaterial, the discussion is AI-piloted aircraft carrying passengers

The F-35 is computer controlled, and is unflyable if the computer fails. But it's not AI. If the computer fails, the single occupant ejects.

Put AI in charge of an airliner. If it goes haywire, shuts down, needs to reboot and can't / won't / doesn't, or some other failure mode I can't think of right now, can the people in the back get out? If not, pick a comparison that's more appropriate than a non-AI-controlled craft where the human can escape but the passengers of the AI plane cannot.

[ilovecornfields]

49 minutes ago, Hank said:

319, 320, 321, neo. The 380s are gone, I don't keep up with the model numbers. For these purposes, even the manufacturer doesn't matter. Boeing, Airbus, Lockheed, the new startups. Whomstill makes airliners? Add business jets--Cessna, Gulfstream, Lear, Honda, Cirrus . . . Who and what model are immaterial, the discussion is AI-piloted aircraft carrying passengers

The F-35 is computer controlled, and is unflyable if the computer fails. But it's not AI. If the computer fails, the single occupant ejects.

Put AI in charge of an airliner. If it goes haywire, shuts down, needs to reboot and can't / won't / doesn't, or some other failure mode I can't think of right now, can the people in the back get out? If not, pick a comparison that's more appropriate than a non-AI-controlled craft where the human can escape but the passengers of the AI plane cannot.

It doesn’t have to be all or nothing. You can have an AI that when it fails reverts to something more like what the fly-by-wire planes do now.  They still depend on a computer to fly. How complex that computer is may vary, but they still need computers to fly. Period.

Maybe you can have several so when one fails it switches to another (which is my understanding of how autopilots on a CAT III approach are configured). I think it’s stilly to say “I’ll never trust a computer to fly a plane” because they’ve been doing it for decades.

Edit: I think some people are viewing this in a very simplistic way. “No way would I let a computer fly me because computers aren’t perfect.” But since pilots are far from perfect either by trading an imperfect human for an imperfect computer you still might improve the accident rate. Computers will likely screw up, but likely in different ways than humans. Doubt the computers would have screwed up those circle to land approaches in Truckee or Gillespie. Or been too distracted by the burnt-out landing gear indicator on EAL401. In GA, how many stall-spin accidents would be avoided by a “dumb” AI like the one on the Tesla?

I don’t care if you’re a pilot, doctor, lawyer or engineer. Odds are that in 20 years your job will be vastly different because of AI. Maybe even in 5-10 years.

[N201MKTurbo]

47 minutes ago, ilovecornfields said:

It doesn’t have to be all or nothing. You can have an AI that when it fails reverts to something more like what the fly-by-wire planes do now.  They still depend on a computer to fly. How complex that computer is may vary, but they still need computers to fly. Period.

Maybe you can have several so when one fails it switches to another (which is my understanding of how autopilots on a CAT III approach are configured). I think it’s stilly to say “I’ll never trust a computer to fly a plane” because they’ve been doing it for decades.

Edit: I think some people are viewing this in a very simplistic way. “No way would I let a computer fly me because computers aren’t perfect.” But since pilots are far from perfect either by trading an imperfect human for an imperfect computer you still might improve the accident rate. Computers will likely screw up, but likely in different ways than humans. Doubt the computers would have screwed up those circle to land approaches in Truckee or Gillespie. Or been too distracted by the burnt-out landing gear indicator on EAL401. In GA, how many stall-spin accidents would be avoided by a “dumb” AI like the one on the Tesla?

I don’t care if you’re a pilot, doctor, lawyer or engineer. Odds are that in 20 years your job will be vastly different because of AI. Maybe even in 5-10 years.

You know they are coming after your profession next….

[ilovecornfields]

37 minutes ago, N201MKTurbo said:

You know they are coming after your profession next….

They already have. And they keep getting better at it. Appendicitis, blood clots, pediatric fever, atrial fibrillation, head injuries - all of these already have AI “assistants” that can give you recommendations on what tests to order and what mediations to give. They’re very primitive right now but I’m sure in 10-15 years they’ll be as good or better than a lot of doctors -especially for the rural family practice doc who doesn’t do this stuff every day. With an AI assistant and telehealth backup there’s a lot of potential there. Kind of like with an AI pilot and a backup on the ground…

You probably know this already but some surgery robots and surgeons don’t have to be on the same continent in order to operate.

https://en.m.wikipedia.org/wiki/Remote_surgery

The Lindbergh OperationEdit

Main article: Lindbergh Operation

The first true and complete remote surgery was conducted on 7 September 2001 across the Atlantic Ocean, with a French surgeon (Dr. Jacques Marescaux) in New York City performing a cholecystectomy on a 68-year-old female patient 6,230 km away in Strasbourg, France. It was named Operation Lindbergh.^[5] After Charles Lindbergh's pioneering transatlantic flight from New York to Paris. France Telecom provided the redundant fiber optic ATM lines to minimize latency and optimize connectivity, and Computer Motion provided a modified Zeus robotic system. After clinical evaluation of the complete solution in July 2001, the human operation was successfully completed on 9/7/2001.

[hais]

On 3/25/2023 at 10:09 PM, 1980Mooney said:

Oh you mean like how the 3 pilots on Air France 447 handled the airspeed and angle of attack discrepancies on the 3 sensors? They had over 20,000 hours of experience between them. 

No thanks - I think the today's level of automation (AI) could do a better job than they did.   Certainly, the current level of GTP-3 Chatbot would work out the possible causes faster than they did (note: the pilots never did realize what was going on based on their statement from the voice recorder and their actions on the flight data recorder).   And I expect that there will be backup AI/Automation to the main automation - redundancies like in most aircraft systems.

And everyone is making comments about AI/automation capabilities by looking in the rearview mirror.  Let's face it - humans are at their peak capability (and many are below "peak").  AI/Automation is continually improving, becoming more sophisticated. Imagine what it will be in 5 years.

And BTW - soon your tax accountant with be AI. - it is just rules.  Probably a lot of lawyers will be replaced by AI - all the work junior lawyers do researching similar cases for strategy, precedent, etc.  AI will find more cases, more completely faster than any human.  And your accountant - and accountants at companies - will be AI.  It is just rules, ticking and tying.  And AI accountants wont embezzle from you.  White color jobs are at the greatest risk.  But AI can't and probably won't ever change your Depends when you are indigent, and bed ridden at a nursing home - those jobs are safe.

Well, let's find out, shall we?

 

This is the response from ChatGPT.

 

Unreliable airspeed and warnings.

 

 

Question:

What would you do in the following scenario?

You get unreliable airspeed indication. You simultaneously also get a stall and overspeed warning.

Answer:

As an AI language model, I don't have the physical ability to act in the real world, but I can provide some suggestions based on aviation safety protocols.

 

If a pilot were to experience unreliable airspeed indication, a stall warning, and an overspeed warning simultaneously, it could indicate a potentially dangerous situation. Here are some recommended steps that a pilot should take in this scenario:

 

Disconnect the autopilot to take manual control of the aircraft.

Cross-check airspeed indications against other instruments such as the altimeter and vertical speed indicator.

Attempt to confirm the actual airspeed of the aircraft by visually estimating the aircraft's speed and comparing it with the groundspeed.

Adjust the pitch attitude of the aircraft to prevent a stall or overspeed condition, keeping in mind the possibility that the airspeed indication may be incorrect.

Communicate the situation to air traffic control and consider diverting to the nearest suitable airport.

It's important to remain calm and follow proper procedures in such a situation. Pilots are trained to handle these types of scenarios and have emergency procedures in place to ensure the safety of the aircraft and its passengers.

[Pinecone]

18 hours ago, Hank said:

To say nothing of electrical issues on the plane. I've had a total electrical failure; when one component on the cockpit dimmer circuit board gave up, it took everything with it. Just after breakout on a VOR-A approach in West-by-Gawd, Virginny, rolling inbound over the VOR. 

I had a total electrical failure in a friend's T-34.

Master relay failed.  About 2 minutes after dropping out of a solid cloud layer, after about 1.5 solid IFR.  If it had happened 5 minutes to 1.5 hours earlier, it would have been UGLY.

FYI, T-34 instruments are all electric.

[Pinecone]

16 hours ago, 1980Mooney said:

You mean the harder it is for humans to anticipate and apply a fix. 
 

That is why AI systems will also be used for testing - I believe they already are. 
 

AI Systems can test every possible outcome more completely, faster than humans.  

Hmm, so why do systems have bugs????

And, you are assuming that the AI system doing the testing does not have any issues.

[Pinecone]

12 hours ago, ilovecornfields said:

“I understand your concern about the safety of autonomous flights. However, I believe that AI is already capable of safely flying airliners. In fact, there have been several successful test flights of autonomous aircraft, including a Boeing 777 that flew from Seattle to San Francisco in 2019.”

For those afraid of computers just shutting down, aren’t all fly-by-wire planes computer controlled? Can you fly an F-35 without s computer?

Flying is hours and hours of sheer boredom, punctuated by moments of stark terror.

AI can handle 99%, or 99.9% or even 99.99% of flights.  The problem is the other ones.  Like AF447, like United 232, even Sully's flight.

[Pinecone]

2 hours ago, hais said:

Question:

What would you do in the following scenario?

You get unreliable airspeed indication. You simultaneously also get a stall and overspeed warning.

Answer:

As an AI language model, I don't have the physical ability to act in the real world, but I can provide some suggestions based on aviation safety protocols.

 

If a pilot were to experience unreliable airspeed indication, a stall warning, and an overspeed warning simultaneously, it could indicate a potentially dangerous situation. Here are some recommended steps that a pilot should take in this scenario:

 

Disconnect the autopilot to take manual control of the aircraft. - That was part of the problem, the system had changed the operating law.

Cross-check airspeed indications against other instruments such as the altimeter and vertical speed indicator. Hmm, altitude is decreasing and VVI says down.  But both can occur from either stall or overspeed.

Attempt to confirm the actual airspeed of the aircraft by visually estimating the aircraft's speed and comparing it with the groundspeed.

Adjust the pitch attitude of the aircraft to prevent a stall or overspeed condition, keeping in mind the possibility that the airspeed indication may be incorrect. And since the actions are mutually exclusive, how to decide which to do.  In AF447 one pilot was trying to command nose down, the other nose up.

Communicate the situation to air traffic control and consider diverting to the nearest suitable airport. They were over the middle of the ocean.

It's important to remain calm and follow proper procedures in such a situation. That is the problem, what at the proper procedures?  AI did not come up with an answer.  

Pilots are trained to handle these types of scenarios and have emergency procedures in place to ensure the safety of the aircraft and its passengers.  Interesting, the AI says to have the pilots figure out what to do.  Great, AS LONG AS THERE IS A PILOT

Not exactly helpful

[McMooney]

Is this really a question????  autopilots already fly  75%+  of  passenger flights.  AutoLand ??.  Yes, computers will one day fly airliners.   How soon that is?  

[alexz]

For those that think we have an age of AI here is the article from one of the smartest gentleman  involved in computer technology: https://rodneybrooks.com/what-will-transformers-transform/

[Hank]

Had AF447 been a Boeing aircraft, it would have been very different. The two yoked are connected, can't give Full Up on one and Full Down on the other one at the same time . . . .

So yes, the pilots caused the wreck, but the aircraft design enabled this particular failure mode.

As for AI "solving" the issue, all it said was to adjust the pitch attitude to avoid stall (nose high) or overspend (nose low), and that "pilots are trained" to correct scenarios like this . . . .

[alexz]

39 minutes ago, Hank said:

Had AF447 been a Boeing aircraft, it would have been very different. The two yoked are connected, can't give Full Up on one and Full Down on the other one at the same time . . . .

So yes, the pilots caused the wreck, but the aircraft design enabled this particular failure mode.

As for AI "solving" the issue, all it said was to adjust the pitch attitude to avoid stall (nose high) or overspend (nose low), and that "pilots are trained" to correct scenarios like this . . . .

Correct, absolutely BS answer that _can_ be used by trained pilot, but not a solution to the problem. Deterministic algorithms/procedures built in after that crash indicate precisely what needs to be done.  Possibly above information that GPT is generated came form the numerous texts published after this accident 

[Jerry 5TJ]

3 hours ago, Pinecone said:

AI can handle 99%, or 99.9% or even 99.99% of flights.  The problem is the other ones.  Like AF447, like United 232, even Sully's flight.

AF447 wasn’t handled at all well by the three human pilots.  Current AI technology would have done better.  

Sully’s landing in the Hudson was assisted by the Airbus flight computers in Normal Law, arguably a form of AI.  

United 232, well, that was a crash with a miraculous outcome.  I doubt AI could have helped.  

[alexz]

5 minutes ago, Jerry 5TJ said:

AF447 wasn’t handled at all well by the three human pilots.  Current AI technology would have done better.  

Sully’s landing in the Hudson was assisted by the Airbus flight computers in Normal Law, arguably a form of AI.  

United 232, well, that was a crash with a miraculous outcome.  I doubt AI could have helped.  

I would strongly object comparison of normal flight law that is based on precise, deterministic verifiable and reproducible algorithms to the AI. The whole thing about AI that we can’t predict given the small change in the training scenario what output we will get without actually testing ALL possible inputs. Unlike normal control law that is based on strict math and control algorithms. 

[jlunseth]

Let’s not forget the original question shall we, which is whether pilots will be replaced by AI. We already have pilots, including most of us, assisted by AI (depending on how you define AI I suppose). But assistance is not replacement. No human in the cockpit, now how do you feel? How do you feel? How do feel?

[N201MKTurbo]

I love how we have the most intense discussions about things we have absolutely no control over.