Then why are so many critics worrying that computers could wipe out humanity? And what does intelligence factor into it. Computers are good at rapidly performing calculations. They don't really need to be smarter than their creator..
United 767-300 Has Fuselage Buckled by Hard Landing at IAH July 29, 2023
- Thread starter john_pifer
- Start date
I see…
So on one hand, we have the President of the United States outlining a goal, inspiring the American people, and gaining Congressional backing with an enormous Federal budget - and on the other, we have you, posting on BITOG, making a claim.
These events seem comparable to you?
Then, clearly, they’re not going to be flying airplanes if they’re less intelligent than the average person.
Four words about pilots + AI: Boeing 737 Max 8. In the two fatal crashes, the pilots fought to override the programming that crashed the jetliners. If the pilots could have overridden the computer, the two flights probably would not have crashed. And the Max 8 software wasn't even technically AI.
That is an excellent point.
The computer software was confused, and forced the nose down, over-riding the pilots.
Hundreds died as a result.
MolaKule
Staff member
I think you are referring to fictional accounts in books and movies, which, while they may make for interesting plots and sells books and movies, not all the facets (science) of AI and robots are presented to the reader or viewer. I really don't see this response as being germane to the topic at hand; that of AI verses pilots in aviation, which was a drift away from the original topic of how this 'buckle' could happen.
Because one of the factors in Human Intelligence is that of context and being able to react to that context, as in rapidly changing, dynamic situations. As Astro14 commented regarding CHAT GPT's answer: "The reason this airplane has a buckled fuselage is that the nose wheel came down too hard. AI wants to LAND it on the nosewheel. Billions of calculations per second are meaningless without the contextual understanding of what AI is calculating. It leads to horribly wrong answers, like this one." Post #76.
Again, you're confusing megaflops as being on par with human intelligence; they are not comparable.
But isn't this what you were implying in previous posts, that AI would become more intelligent than the programmer, the Human?
Last edited:
No way am I getting in a self flying plane.
In the case of the Max, there was also a third crew that saved the day and used the electric trim button on the yoke very aggressively to correct the repeated bursts of nose down trim the MCAS system was delivering.
Had any of the three crews that encountered the MCAS failures truly understood what was going on, they would have trimmed to neutral stick force and then shut of the electric trim off which would have also negated any possibility of any further MCAS trimming. FWIU, the manual trim wheels are fine as long as the aircraft is not allowed to get too far away from neutral stick forces.
In the case of the Lion Air crew that saved their ship and their passengers, they demonstrated an ability to compensate for a potentially deadly event that one would not expect from AI.
IOW, the pilots were perfectly able to override the computer. They just didn't have sufficient understanding of what the problem was and Boeing wasn't entirely forthcoming with information as to an MCAS failure and the actions then to be taken.
Last edited:
UM Ford Center for Autonomous Vehicles (FCAV)
The University of Michigan and Ford Motor Company have partnered form the UM & Ford Center for Autonomous Vehicles (FCAV) to accelerate autonomous vehicle research and explore the future of driving through safer and more capable vehicles. We are researching perception, control, and planing for l
fcav.engin.umich.edu
Fine tuning the autonomous driving and pedestrian recognition is all being worked on. is it perfect right now, no. But it has come a long way in the last 10 years.
For planes they will implement something similar to V2V and V2I technology to know where everything is in relation to everything else. sensors will also need to be deployed as "waypoints" where the vehicle checks into.
MolaKule
Staff member
The problem with this Pie in the Sky thinking for aviation is this:
"Vehicle-to-infrastructure (V2I or v2i) is a communication model that allows vehicles to share information with the components that support a country's highway system. Such components include overhead RFID readers and cameras, traffic lights, lane markers, streetlights, signage and parking meters. V2I communication is typically wireless and bi-directional: data from infrastructure components can be delivered to the vehicle over an ad hoc network and vice versa. Similar to vehicle-to-vehicle (V2V) communication, V2I uses dedicated short range communication (DSRC) frequencies to tranfer data..."
None of the above in bold is available for aviation. How could this possibly be implemented for aviation?
None of this seems to take into account a critical element, that of "Human Factors."
I'm not taking any sides here, but it does seem like you may have dodged the question in lieu of an attack or at least a barb. Clearly his knowledge on the topic is lacking (as is mine to be clear) but he did raise an interesting question...
Replace the government with dozens of large corporations with deep pockets, the percieved need to implement AI pilots for increased safety and the creativity of people and I can see it having potential for success, even if limited. To his point, how many things were considered absolute lunacy only to become common place a few to many years later?
All that said, I'm in the camp that would much rather see trained and accomplished humans like yourself in the front of a plane making decisions than AI. I agree that our ability to feel, perieve and think outside the box is what we need in these applications. Bigger picture, I'm not a fan of large scale AI implementation - if personally feel that the possibility for bad outweighs the good.
Look, I’m happy to talk about Aviation.
Where I get frustrated is that every non-Pilot seems to think they understand what the job is. Well they clearly don’t. Half of the guys in this forum think they could land an airliner with zero prior training, which means half of the guys in this forum are utterly and completely clueless when it comes to understanding Aviation.
Without understanding everything that the job entails, you’re simply not equipped to figure out what is necessary to replace the pilot. That should be simple enough to understand, but it does not prevent the clueless from posting their opinion.
Boeing has been working on it for a long time. The complete, abject failure of their MCAS system to make the airplane safer, when, in fact it did the opposite and killed hundreds, shows you how much progress has been made by big money and big corporations.
At the very best, the increase in computing power is simply going to be deployed as an aid to the Pilot. Just as fly by wire systems have to date. Just remember, MCAS was intended as an aid to the pilot, but some software engineers decided it needed more authority, so they added it, without checking with flight test, without checking with pilots.
The result was the two 737 MAX crashes.
Look, the Air Force has been operating drones for a long time, but they’re remotely piloted, they’re not run by AI, and they crash at a horrible rate. Nobody cares cause they only cost a couple million bucks and nobody gets killed. So, the predator UAV as an analog for remotely piloted airliners is ridiculous.
The Navy has spent years, over a decade, trying to integrate the MQ 25 drone into its carrier operations. Without getting classified, there are severe limitations to what the drone can do. Sure it’s autonomous. That doesn’t mean it’s ready to operate in a commercial environment, in crowded air space. The Navy is barely able to get the thing to take off and land at the same runway every time, so, heaven help them if you want to operate at commercial airports, where the weather, approach procedure and runway changes constantly. Autonomous airplanes are so very far from being viable technology at the standard needed for passenger safety.
So, no, we’re not gonna put AI in charge of passenger airplanes anytime soon, because the very simplest of automated systems have killed people. People do not understand how conservative the regulatory environment, the engineering culture, people, and companies in Commercial Aviation truly are. No one is adopting anything like AI to pilot an airliner anytime soon, despite the fantasies and uninformed prognostications to the contrary.
For me the risks are just too great.
A self driving car that has a system failure will at best pull its self off the road safely and at worst kill the occupants, max 7 people I would think. This is bad enough.
Now take a plane, how many people do the biggest ones carry? Google says it’s an Airbus with 853 seats. So add crew onto that. Got to be at least 10 people so you’ve got 863 people on a plane.
Let’s assume some clever person manages to get the plane to takeoff fly to a destination and land all by its self. Now ask yourself what happens when part of that system fails. The plane cannot “pull off the road” and it cannot now land Because the AI system has failed. What are you going to do? Put a full second system on the plane that can take over? A second everything, down to every wire and sensor?
What happens if a sensor fails but keeps sending a bogus value?the AI doesn’t know because the value is classed as plausible. The AI relies on this value to land the plane so it does not make a safe landing.
What more needs to be said.
As I said previously, I will never get on a self flying plane. Not ever or for nobody.
I wouldn’t trust a self driving car and I would want a fall back control system.
Also for the record. I wouldn’t have a clue how to fly or land a plane.
While I'm not a pilot, I did pass the written exam with a 93. I'm not totally clueless as you would seem to think.
So, you took the written for a private pilot?
That makes you more educated than the general populace, but this may be where the misunderstanding comes in: You think you know more than you actually do.
You have no flight experience.
You haven’t done a commercial, multi engine, or instrument rating. You don’t have an ATP.
You took the first step, but there are many thousand more. You need to have those behind you before you can honestly, accurately, discuss what it means to be an airline pilot.
Don’t forget, that’s what we’re talking about here. Airlines, airliners, and airline pilots. This thread is about an airliner that was dinged up by an airline pilot in a passenger operation.
MolaKule
Staff member
I would encourage you to contact your nearest flight training center and start your flight training if feasible.
Unfortunately even the most experienced pilots sometimes make a mistake.
The public thinks it’s easy to land a commercial airliner because they are dumb dumbs.
This week I traveled on a 737, CRJ-900 and on the way home E175.
E175 is my favorite.
The public thinks it’s easy to land a commercial airliner because they are dumb dumbs.
This week I traveled on a 737, CRJ-900 and on the way home E175.
E175 is my favorite.
Cars were a lot more accessible back then as they are now both for entertainment and income which is why I never pursued it then and have no interest now. Doubt any airline would hire older new pilots anyway.
When I was a full-time flight instructor, I found that there was little correlation to having already passed the written exam and how well people could learn to fly the airplane. People who passed the written exams (private, instrument, commercial, ATP) before they started the flying portion of the certificate or rating was highly correlated to how serious they were about obtaining the certificate or rating. People who would just show up to fly and had done no work prior were unlikely to finish.
GON
$100 Site Donor 2024
Yes, everyone loves four across seating (E175) to six across seating (737). Also, less PAX on a jet meaning less stress loading and loading. And the E175 limited overhead space reduces carry ons.
Wonder if you really like the E175 over the 737, but might you really like a passenger jet with less PAX capacity for the boarding, comfort, and deboarding benefits.