Fighter pilots report that the pilot appeared incapacitated.
The plane made a 180 roughly 100 mi NE of DC.
Lots of riveting videos of DC residents hearing sonic booms.
The media says lots of words before mentioning pilot incapacitation.
Biz jet overflies DC, Crashes in VA mountains
Moderators: FrankM, el, Dmmoore
Biz jet overflies DC, Crashes in VA mountains
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.
Re: Biz jet overflies DC, Crashes in VA mountains
Let the conspiracy theories begin: https://www.thedailybeast.com/crashed-p ... D=ref_fark
HR consultant, Yoyodyne Propulsion Systems, Inc.
Re: Biz jet overflies DC, Crashes in VA mountains
The 180 is strange- I ass-ume commanded. Why not a descent.
Bummed that it had passengers.
Shouldn’t pressurization warnings help?
We need automatic expeditious descent routines in fancy auto pilots.
/spam
Bummed that it had passengers.
Shouldn’t pressurization warnings help?
We need automatic expeditious descent routines in fancy auto pilots.
/spam
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.
Re: Biz jet overflies DC, Crashes in VA mountains
Fixed.The 180 is strange- I ass-ume commanded. Why not a descent.
Bummed that it had passengers.
Shouldn’t pressurization warnings help?
We need mandatory 2-pilots crew for anything Piper Cub or bigger.
/spam
Re: Biz jet overflies DC, Crashes in VA mountains
…Except as we saw with Helios, a 6-person crew wasn’t enough for insidious depressurization… one of the six tried gallantly.
We need mandatory 2-pilots crew for anything Piper Cub or bigger.
Fixed.
I guess we need someone always on 02 and to train FAs to operate pressurization and totally automated descent mechanisms and ballistic parachutes…
And ban FL410, or maybe anything >18,000
Thus, I repeat “ /spam” and interesting that the thing turned around…
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.
Re: Biz jet overflies DC, Crashes in VA mountains
It's okay... soon ChatGPT will be flying all the airplanes and then we'll be safe. 
HR consultant, Yoyodyne Propulsion Systems, Inc.
Re: Biz jet overflies DC, Crashes in VA mountains
Good humour includes a dose of truth...It's okay... soon ChatGPT will be flying all the airplanes and then we'll be safe.
...a nice big dose here...
...not_ChatGPT by name, but...
...No 02 required...
Then again, the squid takeover of the universe may precede this and eliminate said need.
And we have tended to say- "sure, let the damn computer monitor EVERYTHING, as long as we have an over ride...although Gabriel has been hanging out with Evan too much.
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.
Re: Biz jet overflies DC, Crashes in VA mountains
I don't know if you realized but I was being sarcastic with what is being discussed somewhere else.…Except as we saw with Helios, a 6-person crew wasn’t enough for insidious depressurization… one of the six tried gallantly.
We need mandatory 2-pilots crew for anything Piper Cub or bigger.
Fixed.
Yeah, and not getting along...although Gabriel has been hanging out with Evan too much.
Re: Biz jet overflies DC, Crashes in VA mountains
I missed the sarcasm- you tend to be serious AND over there has been even more did dieded than here.I don't know if you realized but I was being sarcastic with what is being discussed somewhere else.
Yeah, and not getting along...although Gabriel has been hanging out with Evan too much.
It figures that Evanie would spout off a poorly-founded recommendation for more stuff and feel very firm about said reccomendation.
There is some decent entertainment in that thread.
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.
Re: Biz jet overflies DC, Crashes in VA mountains
ChatGPT jokes aside, this does seem like one of those places where automation could help. A computicator could monitor cabin pressure, and if it becomes clear that the cabin altitude has become excessive and the plane is not in a steep descent, initiate said descent to a safe altitude... keeping terrain clearance in mind.
It could be part of a system such as is already being installed in some airplanes where there's a large red button on the instrument panel labeled "HALP! The pilot(s) seem to have become incapacitated and I'm just some person who can't fly but likes to roast pilots on intranet fora but I'd really like not to did_died today so Hal, could you maybe land the airplane?"
It could be part of a system such as is already being installed in some airplanes where there's a large red button on the instrument panel labeled "HALP! The pilot(s) seem to have become incapacitated and I'm just some person who can't fly but likes to roast pilots on intranet fora but I'd really like not to did_died today so Hal, could you maybe land the airplane?"
HR consultant, Yoyodyne Propulsion Systems, Inc.
Re: Biz jet overflies DC, Crashes in VA mountains
The sky is the limit (pun intended).ChatGPT jokes aside…
There’s already that dead pilot auto land thing for smaller planes.
I can imagine a computer ATC doing a lot better job at not_clearing SWA for takeoff with ITS on “short” final, and etc. x a large number.
My questions:
I’m sure our real pilots can give us scenarios where automation CAUSES problems…(I keep replacing 02 SENSORS on the car while automixture performs OK). And good ole MCAS and burnt out landing gear bulbs…
Will this be one grand AI QRH overlord, or three systems or five or 25?
Regardless, I ass-ume that lots of stuff is coming…If HAL can fly flyboys plane, I can see him losing his right seat transponder reply monitor person.
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.
Re: Biz jet overflies DC, Crashes in VA mountains
I think to some degree the "automation causing problems" thing could be alleviated if there was clear annunciation to the pilot(s) what the automation was doing. For example with MCAS: "MCAS pushing down relentlessly" flashing prominently on some display.
I do know one problem with that is "crimes of omission"... for example, autothrottles not increasing power when the pilots expect it to, because of being in the wrong mode during landing.
But on the other other other hand, the above at least sometimes could be mitigated by more automation. For example there have been multiple accidents caused by planes getting too slow on an otherwise good approach to land (meaning reasonably stable). It should not be to hard to add a little code to the autothrottle to detect when the plane is in approach configuration, and at the least make a warning that "SPEED IS LOW", or better still, add power.
I do know one problem with that is "crimes of omission"... for example, autothrottles not increasing power when the pilots expect it to, because of being in the wrong mode during landing.
But on the other other other hand, the above at least sometimes could be mitigated by more automation. For example there have been multiple accidents caused by planes getting too slow on an otherwise good approach to land (meaning reasonably stable). It should not be to hard to add a little code to the autothrottle to detect when the plane is in approach configuration, and at the least make a warning that "SPEED IS LOW", or better still, add power.
HR consultant, Yoyodyne Propulsion Systems, Inc.
Pilot was Retired WN Captain
The article has a number of photographs - as it is behind a paywall, here is an excerpt about the crash investigation and the pilot.
National Transportation Safety Board investigators spent Tuesday at a Virginia crash site, combing through the wreckage of a plane linked to a sonic boom heard across Washington as a picture emerged of the pilot as an experienced, safety-focused former airline captain.
The location of the crash, the destruction of the plane and questions about whether the Cessna Citation had an operating black box highlight the arduous job of NTSB investigators tasked with determining what happened.
Investigators are encountering a challenging scene in difficult-to-access woods in the western Virginia mountains. The plane was destroyed in the crash, leaving it no longer identifiable as an aircraft, lead investigator Adam Gerhardt said near the scene. Recovered wreckage will probably be removed by helicopter. The NTSB didn’t know Tuesday whether the plane was operating with voice and data recorders that could help determine the cause of the crash.
“This is a rather extreme example of an airplane that impacted terrain,” Gerhardt said. “It’s already a challenging process and it makes it that much more challenging for us, but we will be here for as long as it takes.”
Hefner was a retired Southwest Airlines captain and former member of its pilots union’s board of directors, the group confirmed. He had recently obtained the highest-level Federal Aviation Administration medical certificate and was rated to fly Boeing 737s, among other planes.
“Jeff was a defender of his fellow pilots’ safety, careers, and family,” the Southwest Airlines Pilots Association said in a statement. “We offer our deepest condolences to his wife, his family, and his friends. The aviation community has lost a true champion.”
Florida attorney Dan Newlin said Hefner had flown him at least 100 times over the past five years. He described Hefner as “Mr. Safety.”
“When it came to flying, he was always super serious, super cautious and very focused,” Newlin said. “He knew aviation inside and out. It was his passion.”
Newlin said he spends considerable time sifting through the records of potential pilots for his law firm, adding that his decision to hire Hefner was easy. He pointed to Hefner’s flight history, which he said included 25 years and more than 25,000 flight-hours with Southwest. He also was certified as an aircraft mechanic, Newlin said.
With Hefner as his captain over the years, Newlin said he learned how much Hefner prioritized his wife and three children. Most other pilots, Newlin said, liked to tack on extra days during trips to enjoy new places. Hefner was consistently on the first commercial flight home to see his family.
“That is what I admire most and remember most about Jeff,” Newlin said. “He really put his family first.”
In the air, Hefner knew how to steer taildraggers — planes that have a landing wheel at the back — propjets and B-17s. On the ground, according to his former co-captain, he knew where to find the best aircraft museums and how to speak in detail about radial engines and jet engines.
“I am telling you, he was very involved in aviation,” said Giovanni Atiles Garcia, 49, a pilot who said he had done about 30 flights with Hefner, including as recently as last week. “I know airplanes, but he knew the details.”
Garcia described Hefner as cautious in the air, a sentiment echoed by multiple passengers who have flown with him. When the two flew private planes together, Garcia said, Hefner used standard practices from his days flying commercial planes — which are far more detailed and formal than what is required on smaller aircraft.
Garcia said Hefner invited him to spend the night at his family home in Satellite Beach, Fla. He said Hefner liked to cook steak while his wife prepared vegetables. The couple chatted over dinner alongside their son and two daughters about how much they enjoyed their community, urging Garcia to consider moving to the neighborhood.
The two met up last week to test-fly a plane that had come out of maintenance. Airborne, they ticked through a checklist — confirming the plane was safe to fly.
Garcia said he noticed Hefner’s lock screen on his phone. It was of his wife, smiling in front of a jet.
On Sunday, Hefner was flying Azarian home to New York after a four-day vacation with Rumpel. It’s not clear what happened aboard the aircraft, and the NTSB’s investigation will probably take at least a year to complete, with a preliminary report expected in about three weeks. The investigation will delve into the condition of the plane, its maintenance history, the role of the weather, and Hefner’s pilot records. The plane’s full history on file with the FAA was not immediately available Tuesday.
https://www.washingtonpost.com/transpor ... stigation/
National Transportation Safety Board investigators spent Tuesday at a Virginia crash site, combing through the wreckage of a plane linked to a sonic boom heard across Washington as a picture emerged of the pilot as an experienced, safety-focused former airline captain.
The location of the crash, the destruction of the plane and questions about whether the Cessna Citation had an operating black box highlight the arduous job of NTSB investigators tasked with determining what happened.
Investigators are encountering a challenging scene in difficult-to-access woods in the western Virginia mountains. The plane was destroyed in the crash, leaving it no longer identifiable as an aircraft, lead investigator Adam Gerhardt said near the scene. Recovered wreckage will probably be removed by helicopter. The NTSB didn’t know Tuesday whether the plane was operating with voice and data recorders that could help determine the cause of the crash.
“This is a rather extreme example of an airplane that impacted terrain,” Gerhardt said. “It’s already a challenging process and it makes it that much more challenging for us, but we will be here for as long as it takes.”
Hefner was a retired Southwest Airlines captain and former member of its pilots union’s board of directors, the group confirmed. He had recently obtained the highest-level Federal Aviation Administration medical certificate and was rated to fly Boeing 737s, among other planes.
“Jeff was a defender of his fellow pilots’ safety, careers, and family,” the Southwest Airlines Pilots Association said in a statement. “We offer our deepest condolences to his wife, his family, and his friends. The aviation community has lost a true champion.”
Florida attorney Dan Newlin said Hefner had flown him at least 100 times over the past five years. He described Hefner as “Mr. Safety.”
“When it came to flying, he was always super serious, super cautious and very focused,” Newlin said. “He knew aviation inside and out. It was his passion.”
Newlin said he spends considerable time sifting through the records of potential pilots for his law firm, adding that his decision to hire Hefner was easy. He pointed to Hefner’s flight history, which he said included 25 years and more than 25,000 flight-hours with Southwest. He also was certified as an aircraft mechanic, Newlin said.
With Hefner as his captain over the years, Newlin said he learned how much Hefner prioritized his wife and three children. Most other pilots, Newlin said, liked to tack on extra days during trips to enjoy new places. Hefner was consistently on the first commercial flight home to see his family.
“That is what I admire most and remember most about Jeff,” Newlin said. “He really put his family first.”
In the air, Hefner knew how to steer taildraggers — planes that have a landing wheel at the back — propjets and B-17s. On the ground, according to his former co-captain, he knew where to find the best aircraft museums and how to speak in detail about radial engines and jet engines.
“I am telling you, he was very involved in aviation,” said Giovanni Atiles Garcia, 49, a pilot who said he had done about 30 flights with Hefner, including as recently as last week. “I know airplanes, but he knew the details.”
Garcia described Hefner as cautious in the air, a sentiment echoed by multiple passengers who have flown with him. When the two flew private planes together, Garcia said, Hefner used standard practices from his days flying commercial planes — which are far more detailed and formal than what is required on smaller aircraft.
Garcia said Hefner invited him to spend the night at his family home in Satellite Beach, Fla. He said Hefner liked to cook steak while his wife prepared vegetables. The couple chatted over dinner alongside their son and two daughters about how much they enjoyed their community, urging Garcia to consider moving to the neighborhood.
The two met up last week to test-fly a plane that had come out of maintenance. Airborne, they ticked through a checklist — confirming the plane was safe to fly.
Garcia said he noticed Hefner’s lock screen on his phone. It was of his wife, smiling in front of a jet.
On Sunday, Hefner was flying Azarian home to New York after a four-day vacation with Rumpel. It’s not clear what happened aboard the aircraft, and the NTSB’s investigation will probably take at least a year to complete, with a preliminary report expected in about three weeks. The investigation will delve into the condition of the plane, its maintenance history, the role of the weather, and Hefner’s pilot records. The plane’s full history on file with the FAA was not immediately available Tuesday.
https://www.washingtonpost.com/transpor ... stigation/
Re: Biz jet overflies DC, Crashes in VA mountains
Guys, where have you been living?"Chat GPT"
For single-pilot incapacitation: https://discover.garmin.com/en-US/autonomi/#autoland
For high-altitude depressurization (which my affect more than 1 pilot): https://discover.garmin.com/en-US/autonomi/#edm
For loss of control / spatial disorientation: https://discover.garmin.com/en-US/autonomi/#esp
For loss of control in engine failures in twins: https://discover.garmin.com/en-US/autonomi/#srb
Ok but what if an engine fails and it was a not-twin? https://discover.garmin.com/en-US/autonomi/#sg
Re: Biz jet overflies DC, Crashes in VA mountains
AOPA ASI's take on this (pretty much same as everyone else that hasn't read the final report): https://www.youtube.com/watch?v=SFZlNCMARVc
Assuming hypoxia due to lack of pressurization turns out to be the problem, I have just one question: Why the actual fark would a nice airplane like that not have a cabin pressure warning? I could build a device to do that with three electronic components for about $20. Maybe (or maybe not) on the cheapest pressurized plane on the market I could see omitting that for cost reasons, but on a corporate jet??
Assuming hypoxia due to lack of pressurization turns out to be the problem, I have just one question: Why the actual fark would a nice airplane like that not have a cabin pressure warning? I could build a device to do that with three electronic components for about $20. Maybe (or maybe not) on the cheapest pressurized plane on the market I could see omitting that for cost reasons, but on a corporate jet??
HR consultant, Yoyodyne Propulsion Systems, Inc.
Re: Biz jet overflies DC, Crashes in VA mountains
[Italics not_on]We[Italics still not_on] are aware of all of this...Guys, where have you been living?"Chat GPT"
For single-pilot incapacitation: https://discover.garmin.com/en-US/autonomi/#autoland
For high-altitude depressurization (which my affect more than 1 pilot): https://discover.garmin.com/en-US/autonomi/#edm
For loss of control / spatial disorientation: https://discover.garmin.com/en-US/autonomi/#esp
For loss of control in engine failures in twins: https://discover.garmin.com/en-US/autonomi/#srb
Ok but what if an engine fails and it was a not-twin? https://discover.garmin.com/en-US/autonomi/#sg
BUT
What I think Eric and I are getting at is ONE INTEGRATED OVERALL SYSTEM The architecture could still be separate SYSTEMS but a "single interface mentality" that looks after almost everything.
Hey, Pierre, the airspeed indication looks unreliable, so chill and check the pitot heat and CALMLY determine if the plane MIGHT be flying fat dumb and happy- with EXACTLY these power and pitch settings that Evan would like...
This is highly irregular Dave...turn on the pressurization....Dave...I'm landing the plane Dave...
Hui Theiu Lo...you are losing airspeed AND drifting below the glideslope and the power is idled...is your poor brain over-loaded with acronym-mode on the autothrottles? You might want to pretend this is a 17772 and grab a handfull of power levers...or how about you let me land the plane?
This is MCAS, we are pushing down relentlessly (and want you to KNOW about this, instead of NOT_telling you)...and remember, the damn trim system get's mechanically bound up if you don't fix this soon...Maybe the system LOOKS at the whole situation and doesn't do the relentless push over...
We are living in the present...and while planes are loaded with LOTS of smart automation- Bonin was STILL hit with MULTIPLE warnings that the world was ending...and Bizjets (and occasional 737 pilots) are did hypoxiaded people and falling (not_yet) on elementary schools...
JMO- but a lot MORE stuff is coming, and Eric and I imagine a Chat GPT-like interface and maybe some effort at a grand overarching logic of problem solving for "everything". (Not sure it will do rudder reversals or busted jack screws).
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.
Re: Biz jet overflies DC, Crashes in VA mountains
For the same reason that cars are not equipped with an low-oil-pressure warning.Why the actual fark would a nice airplane like that not have a cabin pressure warning
Re: Biz jet overflies DC, Crashes in VA mountains
I don't think an active Chat GPT-like system will be implemented in aviation any time soon.JMO- but a lot MORE stuff is coming, and Eric and I imagine a Chat GPT-like interface and maybe some effort at a grand overarching logic of problem solving for "everything". (Not sure it will do rudder reversals or busted jack screws).
I do see more integration and a boarder scope of alerting systems and active systems, but they are going to be procedural (responses triggered by a pre-programmed rigid set of criteria), not generative AI figuring it out on-the-spot.
Re: Biz jet overflies DC, Crashes in VA mountains
I assume you're kdding?For the same reason that cars are not equipped with an low-oil-pressure warning.Why the actual fark would a nice airplane like that not have a cabin pressure warning
If that aircraft type does not have a cabin pressure warning system, it would be the highest level of ironing that that's the case yet it has (I assume) a low-oil-pressure warning system for its engines.
HR consultant, Yoyodyne Propulsion Systems, Inc.
Re: Biz jet overflies DC, Crashes in VA mountains
The question "why is something false is true" cannot be answered.I assume you're kdding?For the same reason that cars are not equipped with an low-oil-pressure warning.Why the actual fark would a nice airplane like that not have a cabin pressure warning
If that aircraft type does not have a cabin pressure warning system, it would be the highest level of ironing that that's the case yet it has (I assume) a low-oil-pressure warning system for its engines.
Cars do have a low-oil-pressure warning.
Re: Biz jet overflies DC, Crashes in VA mountains
I know it's hard to get Evan to understand broad concepts...I don't think Eric and I expect ACTUAL Chat GPT to be operating an aeroplanie with real time data acquisition from www.internet.com...they are going to be procedural (responses triggered by a pre-programmed rigid set of criteria), not generative AI figuring it out on-the-spot.
I do think we will subject Engineers (and computers) to big chunks of data on how areoplanies work and how planes have been crashed and some really good decision trees.
Eric and I also favor a system that watches everything and prioritizes and helps. The DEVELOPMENT of this system might include AI, and there might be an internal database...
But again, a minute ago, you were listing multiple systems that will address big problems with pretty impressive and complex inputs...
Eric and my "nice big oversight system" might do a better job of telling Pierre that the pitot tubes are acting up, and that MCAS thinks you need a nose over and that you glide path and power settings and speed are running out of altitude, airspeed and ideas...Jokes aside, it would pull up Evanie's exact power and pitch setting since choosing one you are familiar with is verboten.
The sky is the limit.
I AGREE that I do not see an overly rapid adoption- the industry has a strong "it ain't broke" philosophy...and the whole topic of having a second pilot around who isn't needed 99.999999% of the time....
Conversely, I do see the opportunity to synthesize things, make nice decision trees and help Bonins and Hypoxic dudes and dudes pulling up relentlessly against Boeing executives...
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.
Re: Biz jet overflies DC, Crashes in VA mountains
I didn't mean actual chat GPT. I meant any form of AI, that is a "thing" that doesn't follow a strict algorithm where for every set of inputs ou now (or at least could potentially know) what the output would be, but that instead you "train" based on a large set of data and expect that the system will come up with an output inferred from comparing the present data being measured. Note that most of the times the data will not agree (i.e. the system would have never seen before the exact combination of inputs as it is presented now), but AI is able to come up with a "probably good solution" based on how things worked out in similar situations it was presented with during training. That is what human pilots do.I know it's hard to get Evan to understand broad concepts...I don't think Eric and I expect ACTUAL Chat GPT to be operating an aeroplanie with real time data acquisition from www.internet.com...they are going to be procedural (responses triggered by a pre-programmed rigid set of criteria), not generative AI figuring it out on-the-spot.
I do think we will subject Engineers (and computers) to big chunks of data on how areoplanies work and how planes have been crashed and some really good decision trees.
Eric and I also favor a system that watches everything and prioritizes and helps. The DEVELOPMENT of this system might include AI, and there might be an internal database...
But again, a minute ago, you were listing multiple systems that will address big problems with pretty impressive and complex inputs...
Eric and my "nice big oversight system" might do a better job of telling Pierre that the pitot tubes are acting up, and that MCAS thinks you need a nose over and that you glide path and power settings and speed are running out of altitude, airspeed and ideas...Jokes aside, it would pull up Evanie's exact power and pitch setting since choosing one you are familiar with is verboten.
The sky is the limit.
I AGREE that I do not see an overly rapid adoption- the industry has a strong "it ain't broke" philosophy...and the whole topic of having a second pilot around who isn't needed 99.999999% of the time....
Conversely, I do see the opportunity to synthesize things, make nice decision trees and help Bonins and Hypoxic dudes and dudes pulling up relentlessly against Boeing executives...
I don't think that being used in aviation anytime soon, especially not in anything active. It maybe used in something that assists with information to the pilot (like the pilot asking "what is the nearest suitable airport" and the system needs to analyze the nearby airports taking into account the airplane's landing performance and weather conditions to give an answer. Maybe a mix of AI and hard code can be used also to detect and alert about abnormal situations or conditions, but I don't think that AI is an advantage for that, not at this stage at least.
For active things, like following the UAS memory items after identifying a UAS event, recovering automatically from an upset, descending to a breathable altitude if the pilot doesn't do that after a los of pressurization, or landing the plane automatically when the pilot is not responsive, or actually anything where the computer will act on the plane, I don't see AI being involved anytime soon. But I do see procedural code being used for these cases. That is at least the direction we are clearing going from Airbus not letting you barrel roll the plane, to the EICAS/ECAM system, to the system that I llnked earlier.
I see us doing more of that and in a more integrated way, rather than shifting to AI.
Re: Biz jet overflies DC, Crashes in VA mountains
Gabe:
I didn't mean actual chat GPT. I meant any form of AI, that is a "thing" that doesn't follow a strict algorithm where for every set of inputs ou now (or at least could potentially know) what the output would be,
3BS:
Trust me, we understand, and basically agree.
Gabe:
but that instead you "train" based on a large set of data and expect that the system will come up with an output inferred from comparing the present data being measured. Note that most of the times the data will not agree (i.e. the system would have never seen before the exact combination of inputs as it is presented now), but AI is able to come up with a "probably good solution" based on how things worked out in similar situations it was presented
3BS:
I can see AI technology “farming” airplane safety data IN A LABORATORY SCENARIO that is used to DESIGN AND IMPROVE your “strict alogorithm” above.
Gabriel
[Blah, blah blah]
I don't think that being done anytime soon
I see us doing more of that and in a more integrated way, rather than shifting to AI.
3BS:
Here’s where I want to cuss you for arguing.
AI IS JUST ALGORITHMS that synthesize “information” and DELIVER IT IN PREFERRED COMMUNICATION METHODS, such as synthetic speech.
I see AI TECHNOLOGY AND ARCHITECTURE having a real fit in the cockpit…
Yeah, it won’t be free form AI, BUT THE ABILITY TO RAPIDLY ANALYZE THINGS AND PROVIDE TIMELY INFORMATION AND MANAGE SCENARIOS seems to be at the doorstep.
Don’t forget, Bonin AND HIS LEFT SEAT PARTNER WERE BOTH BAMBOOZLED by the unhappy FBW computer…
I’m thinking that half of the algorithms present in Chat GPT and a modest database (and yes, a CONTROLLED decision tree) would help them, and Helios, and Hui Thieu Lo and etc x a large number (and the recent wife-saver systems)…the TECHNOLOGY is there. And throw in ATC.
It’s not_pure AI, but AI-inspired.
I didn't mean actual chat GPT. I meant any form of AI, that is a "thing" that doesn't follow a strict algorithm where for every set of inputs ou now (or at least could potentially know) what the output would be,
3BS:
Trust me, we understand, and basically agree.
Gabe:
but that instead you "train" based on a large set of data and expect that the system will come up with an output inferred from comparing the present data being measured. Note that most of the times the data will not agree (i.e. the system would have never seen before the exact combination of inputs as it is presented now), but AI is able to come up with a "probably good solution" based on how things worked out in similar situations it was presented
3BS:
I can see AI technology “farming” airplane safety data IN A LABORATORY SCENARIO that is used to DESIGN AND IMPROVE your “strict alogorithm” above.
Gabriel
[Blah, blah blah]
I don't think that being done anytime soon
I see us doing more of that and in a more integrated way, rather than shifting to AI.
3BS:
Here’s where I want to cuss you for arguing.
AI IS JUST ALGORITHMS that synthesize “information” and DELIVER IT IN PREFERRED COMMUNICATION METHODS, such as synthetic speech.
I see AI TECHNOLOGY AND ARCHITECTURE having a real fit in the cockpit…
Yeah, it won’t be free form AI, BUT THE ABILITY TO RAPIDLY ANALYZE THINGS AND PROVIDE TIMELY INFORMATION AND MANAGE SCENARIOS seems to be at the doorstep.
Don’t forget, Bonin AND HIS LEFT SEAT PARTNER WERE BOTH BAMBOOZLED by the unhappy FBW computer…
I’m thinking that half of the algorithms present in Chat GPT and a modest database (and yes, a CONTROLLED decision tree) would help them, and Helios, and Hui Thieu Lo and etc x a large number (and the recent wife-saver systems)…the TECHNOLOGY is there. And throw in ATC.
It’s not_pure AI, but AI-inspired.
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.
…and yes, aeroengineers will phugoid it up
Quote = Wall Street Journal
A Mental-Health Chatbot Named Tessa Went Rogue
Software designed to deliver pre-written replies got generative AI
BY JULIE JARGON
Amental-health chatbot that veered off script— giving diet advice to people seeking help from an eating- disorder group—was programmed with generative AI without the group’s knowledge.
The bot, named Tessa, was the focus of social-media attention last week when users of the National Eating Disorder Association’s website reported the rogue advice. The bot incident illustrates how AI-enabled assistants can deliver unexpected and potentially dangerous results as they become a bigger part of daily life.
Michiel Rauws, chief executive of San Francisco software developer Cass, said that in 2022 his company rolled out an AI component to its chat-bots, and that included Tessa.
Rauws said Cass acted in accordance with the terms of its contract with NEDA. NEDA, which didn’t pay for the service, took Tessa offline last week.
“We were not consulted about that and we did not authorize that,” said NEDA CEO Liz Thompson about the AI upgrade.
AI assistants trained in the language of therapy present an alluring— though risky—option as demand for physical and mental-health care explodes, and many people are untreated because of a global clinician shortage.
“We simply don’t have enough nurses and doctors to provide the level of care we’re used to and we need technology to help solve for that,” said Rich Birhanzel, the global healthcare industry lead at consulting firm Accenture. But using technology to fill in the gaps must be done carefully, he said.
And from the start, AI chatbots have been known to screw up. In one test chat with Microsoft’s OpenAI-powered Bing chatbot, the software said it would like to steal nuclear codes. Google’s version, Bard, provided incorrect information during its first public demo. And lawyers who recently used OpenAI’s ChatGPT to draft court documents cited nonexistent legal cases that the bot had apparently invented.
A closed system
Researchers at several universities, including Washington University School of Medicine and Stanford University School of Medicine, built Tessa as a closed system. It couldn’t go off-script, said Ellen Fitzsimmons-Craft, an associate professor of psychiatry at Washington University School of Medicine, one of the researchers.
The researchers devised a decision tree of answers to questions people might pose about body image, weight and food. The chatbot couldn’t initially generate new answers from ingested information, the way ChatGPT does.
Tessa was tested in a clinical trial, and considered effective once researchers checked in with users six months later. In February 2022, NEDA decided to offer it as a resource for at-risk visitors to its website (though not to people deemed to have eating disorders).
AI assistants trained in the language of therapy present an alluring— though risky—option.
Cass administered the software for NEDA from Tessa’s start, but the AI component was added later in the year.
“In most cases it performed really well and did and said the right things
and helped people get access to care,” Rauws said. He said in the instances when users pointed out flaws, the company was able to fix them in less than an hour.
Rauws said its data set was restricted to authoritative sources. And as a safeguard, he said, the AI-generated answers came with disclaimers.
Thompson said NEDA didn’t know generative AI had been added and that the organization thought the chatbot was still running on the original closed system.
Eating disorders are complex physical and mental disorders, she added, and when talking to people who have them, “every single word matters.”
Thompson said Tessa will remain offline until NEDA and the university researchers who initially created Tessa revalidate all the chatbot’s content.
“We can’t yet trust AI to offer sound mental-health advice,” said Fitzsimmons-Craft.
Tessa’s unsanctioned advice came to light over the Memorial Day weekend, after news articles said NEDA was replacing its human-staffed informational helpline with Tessa. While the organization did close the helpline, Thompson said the idea it was replaced with a bot is incorrect.
The articles drew people to NEDA’s website where many tested Tessa, according to Cass, peppering the bot with questions about weight loss and healthy eating. Some received the dieting recommendations.
Synthetic empathy
Chatbots are developing a knack for discussing therapeutic or medical situations. Researchers at the University of California, San Diego conducted a study in which participants said ChatGPT provided more empathetic answers than doctors.
Adding generative AI to these bots—that is, the ability to go off script—increases the difficulty of auditing such software, because their answers don’t have a clear origin. Bots are trained on massive troves of text, but they don’t regurgitate passages of that text.
For healthcare, the underlying data sources must be vetted and up-to-date, and even then, the bots trained on them must be well crafted and controlled, Accenture’s Birhanzel said. He said he advises clients to have full visibility into the data model of any chatbot they deploy and to thoroughly test it to ensure unwanted answers don’t slip through. “People have to have extreme confidence if they’re going to rely on something that’s giving advice,” he said.
Despite inherent dangers, software bots are likely to proliferate in clinical settings, Birhanzel said. While Thompson said Tessa wasn’t intended to replace its helpline staff, it is the chatbot—not the humans— that will likely return.
“We’re not shutting down technology,” Thompson said. “But we have to be super careful with the people we serve.”
Yes, Gabriel, this is what we are talk about with pure AI vs AI-inspired, but controlled…
A Mental-Health Chatbot Named Tessa Went Rogue
Software designed to deliver pre-written replies got generative AI
BY JULIE JARGON
Amental-health chatbot that veered off script— giving diet advice to people seeking help from an eating- disorder group—was programmed with generative AI without the group’s knowledge.
The bot, named Tessa, was the focus of social-media attention last week when users of the National Eating Disorder Association’s website reported the rogue advice. The bot incident illustrates how AI-enabled assistants can deliver unexpected and potentially dangerous results as they become a bigger part of daily life.
Michiel Rauws, chief executive of San Francisco software developer Cass, said that in 2022 his company rolled out an AI component to its chat-bots, and that included Tessa.
Rauws said Cass acted in accordance with the terms of its contract with NEDA. NEDA, which didn’t pay for the service, took Tessa offline last week.
“We were not consulted about that and we did not authorize that,” said NEDA CEO Liz Thompson about the AI upgrade.
AI assistants trained in the language of therapy present an alluring— though risky—option as demand for physical and mental-health care explodes, and many people are untreated because of a global clinician shortage.
“We simply don’t have enough nurses and doctors to provide the level of care we’re used to and we need technology to help solve for that,” said Rich Birhanzel, the global healthcare industry lead at consulting firm Accenture. But using technology to fill in the gaps must be done carefully, he said.
And from the start, AI chatbots have been known to screw up. In one test chat with Microsoft’s OpenAI-powered Bing chatbot, the software said it would like to steal nuclear codes. Google’s version, Bard, provided incorrect information during its first public demo. And lawyers who recently used OpenAI’s ChatGPT to draft court documents cited nonexistent legal cases that the bot had apparently invented.
A closed system
Researchers at several universities, including Washington University School of Medicine and Stanford University School of Medicine, built Tessa as a closed system. It couldn’t go off-script, said Ellen Fitzsimmons-Craft, an associate professor of psychiatry at Washington University School of Medicine, one of the researchers.
The researchers devised a decision tree of answers to questions people might pose about body image, weight and food. The chatbot couldn’t initially generate new answers from ingested information, the way ChatGPT does.
Tessa was tested in a clinical trial, and considered effective once researchers checked in with users six months later. In February 2022, NEDA decided to offer it as a resource for at-risk visitors to its website (though not to people deemed to have eating disorders).
AI assistants trained in the language of therapy present an alluring— though risky—option.
Cass administered the software for NEDA from Tessa’s start, but the AI component was added later in the year.
“In most cases it performed really well and did and said the right things
and helped people get access to care,” Rauws said. He said in the instances when users pointed out flaws, the company was able to fix them in less than an hour.
Rauws said its data set was restricted to authoritative sources. And as a safeguard, he said, the AI-generated answers came with disclaimers.
Thompson said NEDA didn’t know generative AI had been added and that the organization thought the chatbot was still running on the original closed system.
Eating disorders are complex physical and mental disorders, she added, and when talking to people who have them, “every single word matters.”
Thompson said Tessa will remain offline until NEDA and the university researchers who initially created Tessa revalidate all the chatbot’s content.
“We can’t yet trust AI to offer sound mental-health advice,” said Fitzsimmons-Craft.
Tessa’s unsanctioned advice came to light over the Memorial Day weekend, after news articles said NEDA was replacing its human-staffed informational helpline with Tessa. While the organization did close the helpline, Thompson said the idea it was replaced with a bot is incorrect.
The articles drew people to NEDA’s website where many tested Tessa, according to Cass, peppering the bot with questions about weight loss and healthy eating. Some received the dieting recommendations.
Synthetic empathy
Chatbots are developing a knack for discussing therapeutic or medical situations. Researchers at the University of California, San Diego conducted a study in which participants said ChatGPT provided more empathetic answers than doctors.
Adding generative AI to these bots—that is, the ability to go off script—increases the difficulty of auditing such software, because their answers don’t have a clear origin. Bots are trained on massive troves of text, but they don’t regurgitate passages of that text.
For healthcare, the underlying data sources must be vetted and up-to-date, and even then, the bots trained on them must be well crafted and controlled, Accenture’s Birhanzel said. He said he advises clients to have full visibility into the data model of any chatbot they deploy and to thoroughly test it to ensure unwanted answers don’t slip through. “People have to have extreme confidence if they’re going to rely on something that’s giving advice,” he said.
Despite inherent dangers, software bots are likely to proliferate in clinical settings, Birhanzel said. While Thompson said Tessa wasn’t intended to replace its helpline staff, it is the chatbot—not the humans— that will likely return.
“We’re not shutting down technology,” Thompson said. “But we have to be super careful with the people we serve.”
Yes, Gabriel, this is what we are talk about with pure AI vs AI-inspired, but controlled…
Commercial Pilot, Vandelay Industries, Inc., Plant Nutrient Division.
Who is online
Users browsing this forum: No registered users and 1 guest
