14 minutes ago, swansont said:

Just because something is not personally experienced by you in no way impacts the truth or validity of others’ experiences. And being a statistical argument, it means that you are looking at the total effect of all events.

Strangely, every person I ever talked to who had cancer was alive when I talked to them. Based this experience I can apparently conclude that cancer is rather benign.

3 hours ago, CharonY said:

Strangely, every person I ever talked to who had cancer was alive when I talked to them. Based this experience I can apparently conclude that cancer is rather benign.

Yup. I’ve never had cancer, ergo it’s no big deal

On 8/25/2025 at 4:38 PM, zapatos said:

It's been my experience that talking on the phone vs talking to a passenger seems much different. In trying to decide why it felt that way to me I concluded it is because the person on the phone is not sharing the driving experience with me and can distract me by continuing to talk when I need to concentrate on driving. A person in the car with me recognizes when I am dealing with a situation that requires my full attention, and thus quits talking for a moment.

If you are talking to a passenger and driving, you can say 'be quiet for a moment' if the situation demands it, so you are able to focus on driving or other drivers.

Just found this course on Futurelearn, which may give some insight into your question

https://www.futurelearn.com/courses/justice-fundamental-rights-and-artificial-intelligence

Paul

Microsoft has recently announced that Copilot is for entertainment only, so I think the lawyers are scared about liability as people blame the AI companies for screwups.

“Copilot is for entertainment purposes only,” the company warned. “It can make mistakes, and it may not work as intended. Don’t rely on Copilot for important advice. Use Copilot at your own risk.”

https://techcrunch.com/2026/04/05/copilot-is-for-entertainment-purposes-only-according-to-microsofts-terms-of-service/

AI will probably have to get over the hurdle that lie detectors never could before being accepted in courts. People will likely try and there will be lawsuits challenging it.

Ain't that convenient, though. "Here is a product that we want to use for all kind of things so that we can train it to be better. However, if there is any liability of any sort that is entirely on you".

For sure, that is the strategy of most large companies, but especially tech has been interwoven so strongly with society via things that folks interact daily, such as social media and now AI, so that the reward/harm balance is way out of whack.

7 hours ago, CharonY said:

Ain't that convenient, though. "Here is a product that we want to use for all kind of things so that we can train it to be better. However, if there is any liability of any sort that is entirely on you".

For sure, that is the strategy of most large companies, but especially tech has been interwoven so strongly with society via things that folks interact daily, such as social media and now AI, so that the reward/harm balance is way out of whack.

In both degree and speed...

On 1/20/2026 at 3:14 PM, Gees said: (In response to Swansont's belief that seatbelts save lives.)
Maybe. I have driven all over the United States, in Europe, on the right side of the road and on the other side, in cars, trucks, rental vehicles, and on motorcycles. I have towed cars, trailers, boats, and campers, but I do not wear seatbelts and have not for most of my life. Yet, I am not dead. It is possible that statistically speaking, seatbelts are not good for me. 😁

On 1/20/2026 at 4:29 PM, swansont said:

Wow, what a crappy argument!

I know that you are intelligent. Maybe that is why I am forever giving you more credit than is apparently warranted. I will try to explain my "crappy" argument. The things listed above are not only about me, they are circumstances that have been researched and found to contribute to accidents -- statistically speaking. Driving on different terrains, like a Michigan flatlander driving down Franklin mountain to meet Highway 10 in El Paso, Texas, with snow on the road (which generally causes a pile up of cars). Or crossing the Rockies pulling a camper, especially after turning onto steep side roads. Or driving in Ireland on the other side of the road, and having the steering wheel on the other side of the car, and using the gear shift with the left hand because it is also backward, then trying to avoid the stone walls that closely edge many of the roads. Of course, driving in downtown Rome at night when they turned all the stop lights into blinking yellow lights both ways was interesting. It took a while to figure out how to cross the streets without hitting another car. Of course, towing a car or backing a boat into the water requires additional skills. Riding a motorcycle requires constant vigilance and specific skills. Car rental companies will tell you that changing vehicles requires extra caution. I have done all of these things without accidents. About the only other things that I could do to encourage accidents -- statistically speaking -- is to drive drunk, which I don't do, or to drive for a living, which I have only occasionally done.

Now you can argue that you are not aware of the research and statistical information that applies to the above, so I have to look up each statistic to prove that this is true. But if you say as much, I will not believe you.

On 1/20/2026 at 4:29 PM, swansont said:

Just because something is not personally experienced by you in no way impacts the truth or validity of others’ experiences.

True, and I didn't state otherwise. If you look at my response above, you will see that I stated, "It is possible that statistically speaking, seatbelts are not good for me." You see that word, "me"? On the other hand, if seatbelts are not good for me, then it is possible that they are not good for some others, as well. We will never know because no one is going to collect data that considers this idea. This is much like the problem of the helmet laws for motorcyclists, which required bikers to wear helmets for years and years, even though helmets were killing people. Many people believed that helmets saved lives -- they did not. But finally the laws changed. One of the problems with statistics is the same as the problem that science has -- confirmation bias. You get answers to questions that you ask -- simple truth. If your premise is faulty and you don't ask the right questions, then you get crap.

On 1/20/2026 at 4:29 PM, swansont said:

And being a statistical argument, it means that you are looking at the total effect of all events.

Yes, all events that apply to that subject, and the subject of that argument was me. That is exactly what I did in my "crappy argument" by listing all of the experiences that were part of my driving record, which should have caused accidents -- statistically speaking. But in truth, statistical arguments do not look at the "total effect", they just look to answer the questions asked.

On 1/20/2026 at 3:14 PM, Gees said:

Yes it is a statistical argument, and statistics are made up of facts -- not truths. The reality is that facts, therefore statistics, are easy to cherry pick, manipulate, and cause an invalid perception. Just look at what comes out of Washington if you want evidence of that truth. I have a lot of issues with statistics and don't trust them.

On 1/20/2026 at 4:29 PM, swansont said:

I’m not sure that willful ignorance is something people find persuasive, especially in a science discussion setting

Oh, I don't know about that. You used it to great effect the last time we argued and got three up votes for it.

On 1/20/2026 at 3:14 PM, Gees said:

In the first place, I worked as a temp in many offices in different capacities for years, and watched clerical workers gathering information (statistics) for bosses, insurance purposes, government agencies, etc. Most workers did not give two sh*ts about the information and saw it as an added, unpaid, responsibility. Would some/many of them fudge the information if it made the job easier? Absolutely.

On 1/20/2026 at 4:29 PM, swansont said:

People being lazy and/or dishonest is an indictment of those people, not statistics. If there is fault to be found you can dig into the methodology for that. But a blanket dismissal is not an argument to be taken seriously.

I suspect that your assumption that the workers were "lazy" or "dishonest" is because they are clerical workers. Would you jump to the same conclusions if they were scientists? I think I was about 15 years old when I first began to doubt statistics. I was reading an article about heart disease in a magazine and how it caused so much death. My Mother and Grandmother were at the dining room table enjoying tea, when I told them about the article. I don't remember the terminology that I used, but they looked at me then at each other then burst out laughing. I didn't see what was so damned funny and said so, but every time one of them tried to explain, they looked at each other and burst out laughing again.

Finally my Mother asked, "What happens when you die?", and my Grandmother replied, "Your heart stops.", then they burst into laughter again. Once they got themselves under control, I learned that in the 45+ years between them that they had worked as nurses, they had witnessed many deaths and knew that heart disease/attack was often used as an explanation for death whether it was causal or not. Of course, that was over 60 years ago and medicine has been much improved since, so that does not happen anymore -- or does it?

Twenty years ago, my husband died of cancer. Everyone knows that he died of cancer, but his Death Certificate states that congestive heart failure contributed to his death, which is not really true. I remember when they diagnosed him with congestive heart failure about six months before he died. I was quite surprised as he was no couch potato; he was very active, had no problems with blood pressure or his cholesterol levels. It made no sense to me. It made no sense to one of the younger doctors either, as he noted that the blood work did not support the diagnosis.

My husband was sent home to die and Hospice was ordered for him. The Hospice doctor immediately took him off the cholesterol medication as my husband's bloodwork confirmed that it was not needed. His congestive heart failure was not caused by his blood. It was caused by the large tumor growing out of the main artery by his spine, which caused so much pressure on his spine that he could not feel his feet. After the surgery, when they tried to remove the tumor (but only got part of it), he was so pleased because he could feel his feet for the first time in a year. He died a few months later.

The Hospice doctor was the one who filled out the Death Certificate, so why did he put congestive heart failure on it? He knew that my husband did not suffer from congestive heart failure, as the enlarged heart was due to the tumor; not cholesterol, high blood pressure, or bad diet/living habits. I would like to suggest that he put it on the certificate because of the rules, regulations, and facts that were required with regard to explaining the death. Much like the clerical workers, who were guided by policy and procedure rather than truth.

So the people gathering statistics and the people studying them will add another lie to their knowledge. Was the doctor "lazy" or "dishonest"? He knew full well that the information would imply a false narrative, but he did it anyway. Would you call that "fudging" the truth so that he would not have to complicate the information? Should he have explained that the congestive heart failure was caused by a tumor? Is there a box for that information on the Certificate? Do the people, who are gathering statistics, want that much information? No.

Truth can be simple; truth can be complex. It can be whole and it can be partial; such as, more true than not, or mostly true, or even occasionally true, or true now, or true from a specific perspective. Truth is rarely easy, which is why we need philosophers to help us find it. If it were easy, we could give it to scientists, they could make up an experiment, test it, and we could all know the truth. But the reality is that in order to fit truth into the little box that defines fact, we will have to corrupt truth.

AI is not about intelligence or truth, it is about assimilating information and statistics. Using statistics to make judgements is like gambling, it is playing the odds. If people want to do that they can, but AI does not belong in a Court of law as it has nothing to do with justice or truth.

Gee

On 8/26/2025 at 12:21 AM, TheVat said:

Also I'm a bit puzzled that talking on the phone would be a ticketable offense

Here in Australia it is an offense - hands-free (bluetooth with on-the-steering wheel or verbal switching) is fine but holding the phone isn't.

It does make sense. The danger isn't so much the talking part, but more from locating a ringing phone, looking to see who the caller is, finding the 'answer' option or conversely, the process of making a call out. Those do require a lot of attention and looking at the phone.

>>Considering all the recent discussions about AI lying to us to satisfy its programming, how should we consider evidence of wrongdoing provided by AI ?

For the topic itself I am more concerned for AI as 'expert witness' than for faking of evidence; it could become problematic (more so with news media than courts) but I think identifying fake images or documents may be less so (probably using a form of AI?) than assessing the competence of AI giving expert testimony. I don't know whether how cross examination might work.

One AI 'expert opinion' seems insufficient for a court but the same 'testimony' across several different, independent AI's may be seen as sufficient - whilst differing testimony would see it rejected.

Edited May 16May 16 by Ken Fabian

2 hours ago, Gees said:

Yes, all events that apply to that subject, and the subject of that argument was me. That is exactly what I did in my "crappy argument" by listing all of the experiences that were part of my driving record, which should have caused accidents -- statistically speaking. But in truth, statistical arguments do not look at the "total effect", they just look to answer the questions asked.

A seatbelt doesn’t save lives of people who don’t get into accidents. You only mentioned driving, which means it’s irrelevant to the issue. If you had said you’d been in an accident and didn’t die despite not wearing a seatbelt, that would be an anecdote, but still irrelevant. The combination of the two make it a crappy argument.

In the US, about half of auto accident fatalities were people not wearing seatbelts, yet the vast majority wear seatbelts.

If seatbelts did nothing, the odds of dying would be the same whether or not you wore one. When looking at a large sample (not anecdotes) you’d expect fewer fatalities since there are fewer people in that group.

https://www.thezebra.com/resources/research/seat-belt-statistics/

13 hours ago, Gees said:

On 1/20/2026 at 3:14 PM, Gees said: (In response to Swansont's belief that seatbelts save lives.)
Maybe. I have driven all over the United States, in Europe, on the right side of the road and on the other side, in cars, trucks, rental vehicles, and on motorcycles. I have towed cars, trailers, boats, and campers, but I do not wear seatbelts and have not for most of my life. Yet, I am not dead. It is possible that statistically speaking, seatbelts are not good for me. 😁

I know that you are intelligent. Maybe that is why I am forever giving you more credit than is apparently warranted. I will try to explain my "crappy" argument. The things listed above are not only about me, they are circumstances that have been researched and found to contribute to accidents -- statistically speaking. Driving on different terrains, like a Michigan flatlander driving down Franklin mountain to meet Highway 10 in El Paso, Texas, with snow on the road (which generally causes a pile up of cars). Or crossing the Rockies pulling a camper, especially after turning onto steep side roads. Or driving in Ireland on the other side of the road, and having the steering wheel on the other side of the car, and using the gear shift with the left hand because it is also backward, then trying to avoid the stone walls that closely edge many of the roads. Of course, driving in downtown Rome at night when they turned all the stop lights into blinking yellow lights both ways was interesting. It took a while to figure out how to cross the streets without hitting another car. Of course, towing a car or backing a boat into the water requires additional skills. Riding a motorcycle requires constant vigilance and specific skills. Car rental companies will tell you that changing vehicles requires extra caution. I have done all of these things without accidents. About the only other things that I could do to encourage accidents -- statistically speaking -- is to drive drunk, which I don't do, or to drive for a living, which I have only occasionally done.

Now you can argue that you are not aware of the research and statistical information that applies to the above, so I have to look up each statistic to prove that this is true. But if you say as much, I will not believe you.

True, and I didn't state otherwise. If you look at my response above, you will see that I stated, "It is possible that statistically speaking, seatbelts are not good for me." You see that word, "me"? On the other hand, if seatbelts are not good for me, then it is possible that they are not good for some others, as well. We will never know because no one is going to collect data that considers this idea. This is much like the problem of the helmet laws for motorcyclists, which required bikers to wear helmets for years and years, even though helmets were killing people. Many people believed that helmets saved lives -- they did not. But finally the laws changed. One of the problems with statistics is the same as the problem that science has -- confirmation bias. You get answers to questions that you ask -- simple truth. If your premise is faulty and you don't ask the right questions, then you get crap.

Yes, all events that apply to that subject, and the subject of that argument was me. That is exactly what I did in my "crappy argument" by listing all of the experiences that were part of my driving record, which should have caused accidents -- statistically speaking. But in truth, statistical arguments do not look at the "total effect", they just look to answer the questions asked.

On 1/20/2026 at 3:14 PM, Gees said:

Yes it is a statistical argument, and statistics are made up of facts -- not truths. The reality is that facts, therefore statistics, are easy to cherry pick, manipulate, and cause an invalid perception. Just look at what comes out of Washington if you want evidence of that truth. I have a lot of issues with statistics and don't trust them.

Oh, I don't know about that. You used it to great effect the last time we argued and got three up votes for it.

On 1/20/2026 at 3:14 PM, Gees said:

In the first place, I worked as a temp in many offices in different capacities for years, and watched clerical workers gathering information (statistics) for bosses, insurance purposes, government agencies, etc. Most workers did not give two sh*ts about the information and saw it as an added, unpaid, responsibility. Would some/many of them fudge the information if it made the job easier? Absolutely.

I suspect that your assumption that the workers were "lazy" or "dishonest" is because they are clerical workers. Would you jump to the same conclusions if they were scientists? I think I was about 15 years old when I first began to doubt statistics. I was reading an article about heart disease in a magazine and how it caused so much death. My Mother and Grandmother were at the dining room table enjoying tea, when I told them about the article. I don't remember the terminology that I used, but they looked at me then at each other then burst out laughing. I didn't see what was so damned funny and said so, but every time one of them tried to explain, they looked at each other and burst out laughing again.

Finally my Mother asked, "What happens when you die?", and my Grandmother replied, "Your heart stops.", then they burst into laughter again. Once they got themselves under control, I learned that in the 45+ years between them that they had worked as nurses, they had witnessed many deaths and knew that heart disease/attack was often used as an explanation for death whether it was causal or not. Of course, that was over 60 years ago and medicine has been much improved since, so that does not happen anymore -- or does it?

Twenty years ago, my husband died of cancer. Everyone knows that he died of cancer, but his Death Certificate states that congestive heart failure contributed to his death, which is not really true. I remember when they diagnosed him with congestive heart failure about six months before he died. I was quite surprised as he was no couch potato; he was very active, had no problems with blood pressure or his cholesterol levels. It made no sense to me. It made no sense to one of the younger doctors either, as he noted that the blood work did not support the diagnosis.

My husband was sent home to die and Hospice was ordered for him. The Hospice doctor immediately took him off the cholesterol medication as my husband's bloodwork confirmed that it was not needed. His congestive heart failure was not caused by his blood. It was caused by the large tumor growing out of the main artery by his spine, which caused so much pressure on his spine that he could not feel his feet. After the surgery, when they tried to remove the tumor (but only got part of it), he was so pleased because he could feel his feet for the first time in a year. He died a few months later.

The Hospice doctor was the one who filled out the Death Certificate, so why did he put congestive heart failure on it? He knew that my husband did not suffer from congestive heart failure, as the enlarged heart was due to the tumor; not cholesterol, high blood pressure, or bad diet/living habits. I would like to suggest that he put it on the certificate because of the rules, regulations, and facts that were required with regard to explaining the death. Much like the clerical workers, who were guided by policy and procedure rather than truth.

So the people gathering statistics and the people studying them will add another lie to their knowledge. Was the doctor "lazy" or "dishonest"? He knew full well that the information would imply a false narrative, but he did it anyway. Would you call that "fudging" the truth so that he would not have to complicate the information? Should he have explained that the congestive heart failure was caused by a tumor? Is there a box for that information on the Certificate? Do the people, who are gathering statistics, want that much information? No.

Truth can be simple; truth can be complex. It can be whole and it can be partial; such as, more true than not, or mostly true, or even occasionally true, or true now, or true from a specific perspective. Truth is rarely easy, which is why we need philosophers to help us find it. If it were easy, we could give it to scientists, they could make up an experiment, test it, and we could all know the truth. But the reality is that in order to fit truth into the little box that defines fact, we will have to corrupt truth.

AI is not about intelligence or truth, it is about assimilating information and statistics. Using statistics to make judgements is like gambling, it is playing the odds. If people want to do that they can, but AI does not belong in a Court of law as it has nothing to do with justice or truth.

Gee

Thank you for taking an interest in my thread.

10 hours ago, swansont said:

A seatbelt doesn’t save lives of people who don’t get into accidents. You only mentioned driving, which means it’s irrelevant to the issue. If you had said you’d been in an accident and didn’t die despite not wearing a seatbelt, that would be an anecdote, but still irrelevant. The combination of the two make it a crappy argument.

In the US, about half of auto accident fatalities were people not wearing seatbelts, yet the vast majority wear seatbelts.

If seatbelts did nothing, the odds of dying would be the same whether or not you wore one. When looking at a large sample (not anecdotes) you’d expect fewer fatalities since there are fewer people in that group.

https://www.thezebra.com/resources/research/seat-belt-statistics/

I fully understand both your points of view about statistics, but would respectfully remind you that statistics has little or nothing to do with the topic of this thread which is about legal matters.

Before leaving the matter I would like to point out that you are both considering entirely different statistical circumstances. Gees haas a sample size of 1, but swansont is considering very large sample sizes.

The topic asks about the legal responsibility of some form of computer system (AI) driving vehicles. Apparantly nobody in some cases.

This issue of significance has recently come to light in California, resulting in a change in their laws.

BBC News

California to begin ticketing driverless cars that violat...

Under the new rules, police will be able to issue tickets directly to the car's manufacturer when an autonomous vehicle breaks a traffic law.

BBC News

Waymo cars become trapped in Atlanta suburb after glitch

The company said the vehicles, which use AI to drive, had encountered "a routing problem" that kept taking them to the same cul-de-sac.

On 5/16/2026 at 7:20 PM, Ken Fabian said:

>>Considering all the recent discussions about AI lying to us to satisfy its programming, how should we consider evidence of wrongdoing provided by AI ?

For the topic itself I am more concerned for AI as 'expert witness' than for faking of evidence; it could become problematic (more so with news media than courts) but I think identifying fake images or documents may be less so (probably using a form of AI?) than assessing the competence of AI giving expert testimony.

I agree. Right now, AI is not so much a problem in the Courts as it is in making policy and procedure IMO. It is surprising to see how many people do not really differentiate between policy/procedure in business and what is law, and there is a huge difference. All businesses, whether private or government, have policies and procedures that regulate everyday work, and AI is beginning to be the go to expert that defines these policies and procedures. Then media reinforces the belief that AI is expert, which lays the foundation for AI to infiltrate law. A lot of people have explained that AI is being trained and learning to be better at its work, but I think there is more to this. I think we are also being trained, or maybe you could say that we are being groomed, to accept AI as expert.

On 5/16/2026 at 7:20 PM, Ken Fabian said:

I don't know whether how cross examination might work.

This made me smile. Generally speaking, the purpose of cross examination is to ferret out the lies, but AI does not actually lie, so this seems pretty meaningless. I started to envision a Disney type movie where AI would cross examine AI and it was a lot of fun in my mind, but I am not sure how that could really work in reality.

On 5/16/2026 at 7:20 PM, Ken Fabian said:

One AI 'expert opinion' seems insufficient for a court but the same 'testimony' across several different, independent AI's may be seen as sufficient - whilst differing testimony would see it rejected.

In order to come up with different 'testimony' from different AI opinions, I believe that different statistics, or additional information, would have to be fed into some of the AI programs. Otherwise the 'testimony' would be the same.

Have you had the opportunity to see the movie, "Sully"? It is a true story that addresses this AI expert opinion problem. In the movie, Sully, Captain Sullenberger, is a commercial airline pilot with over 40 years of flying experience, who takes off from the airport and within minutes loses his engines. He manages to land his plane in the Hudson river saving all passengers, but losing the plane. Air traffic control wanted him to turn the plane around and return to the airport, and AI confirmed that that was the best solution, but Sully argued that his experience told him that he would never make it and everyone would have died. Tom Hanks plays Captain Sully and Clint Eastwood directs this true story. I remember reading about it in the paper when it happened 20 or so years ago and remember that Captain Sully almost lost his job, but his experience knew something that AI had not considered, which won his case.

Gee

On 5/17/2026 at 6:32 AM, studiot said:

Thank you for taking an interest in my thread.

Although I enjoy philosophy, I rarely post in the Ethics subforum. Thank you for writing an interesting thread.

On 5/17/2026 at 6:32 AM, studiot said:

I fully understand both your points of view about statistics, but would respectfully remind you that statistics has little or nothing to do with the topic of this thread which is about legal matters.

But statistics have everything to do with AI, and how AI affects legal matters is what this thread is about. To my mind, this means that statistics have a great deal to do with this thread and its legal matters.

On 5/17/2026 at 6:32 AM, studiot said:

Before leaving the matter I would like to point out that you are both considering entirely different statistical circumstances.

That would be because I am dealing in truths; whereas, Swansont is dealing in facts. Although facts can help one find truth, fact is not truth. Did you ever hear of a Court of law asking for, "The whole fact and nothing but the facts?"

On 5/17/2026 at 6:32 AM, studiot said:

Gees has a sample size of 1, but swansont is considering very large sample sizes.

If you tell a lie one time or a thousand times, does that make a difference in the veracity of the lie? Does the number make a difference in the veracity of truth? I seem to remember a quote, "If you tell a lie long enough and loud enough, people will begin to believe it." Does that make the lie a truth? Maybe it makes the lie a fact?

On 5/17/2026 at 6:32 AM, studiot said:

The topic asks about the legal responsibility of some form of computer system (AI) driving vehicles. Apparantly nobody in some cases.

Which "legal responsibility"? When you started this thread, you appeared to be considering how a Court would judge AI, but it now appears that you are talking about writing laws. Although both have "legal responsibilities", they are very different considerations and processes. When creating law, one must consider a large populace, and so, they often look to facts and large sample sizes so the law will be best for most of the populace. In a Court of law, it is the singular truths that are being judged about an individual or circumstance. Very different.

No matter which responsibility you are considering, truth is required to consider it. Why? Because wisdom is required and wisdom is an advanced level of truth. It is not an advanced level of facts -- that would be statistics.

On 5/17/2026 at 6:32 AM, studiot said:

This issue of significance has recently come to light in California, resulting in a change in their laws.

That is about control and how we use AI to regulate the citizens; and therefore, what responsibility AI companies have to the citizenry.

Gee

8 hours ago, Gees said:

I agree. Right now, AI is not so much a problem in the Courts as it is in making policy and procedure IMO. It is surprising to see how many people do not really differentiate between policy/procedure in business and what is law, and there is a huge difference. All businesses, whether private or government, have policies and procedures that regulate everyday work, and AI is beginning to be the go to expert that defines these policies and procedures. Then media reinforces the belief that AI is expert, which lays the foundation for AI to infiltrate law. A lot of people have explained that AI is being trained and learning to be better at its work, but I think there is more to this. I think we are also being trained, or maybe you could say that we are being groomed, to accept AI as expert.

This made me smile. Generally speaking, the purpose of cross examination is to ferret out the lies, but AI does not actually lie, so this seems pretty meaningless. I started to envision a Disney type movie where AI would cross examine AI and it was a lot of fun in my mind, but I am not sure how that could really work in reality.

In order to come up with different 'testimony' from different AI opinions, I believe that different statistics, or additional information, would have to be fed into some of the AI programs. Otherwise the 'testimony' would be the same.

Have you had the opportunity to see the movie, "Sully"? It is a true story that addresses this AI expert opinion problem. In the movie, Sully, Captain Sullenberger, is a commercial airline pilot with over 40 years of flying experience, who takes off from the airport and within minutes loses his engines. He manages to land his plane in the Hudson river saving all passengers, but losing the plane. Air traffic control wanted him to turn the plane around and return to the airport, and AI confirmed that that was the best solution, but Sully argued that his experience told him that he would never make it and everyone would have died. Tom Hanks plays Captain Sully and Clint Eastwood directs this true story. I remember reading about it in the paper when it happened 20 or so years ago and remember that Captain Sully almost lost his job, but his experience knew something that AI had not considered, which won his case.

Gee

Although I enjoy philosophy, I rarely post in the Ethics subforum. Thank you for writing an interesting thread.

But statistics have everything to do with AI, and how AI affects legal matters is what this thread is about. To my mind, this means that statistics have a great deal to do with this thread and its legal matters.

That would be because I am dealing in truths; whereas, Swansont is dealing in facts. Although facts can help one find truth, fact is not truth. Did you ever hear of a Court of law asking for, "The whole fact and nothing but the facts?"

If you tell a lie one time or a thousand times, does that make a difference in the veracity of the lie? Does the number make a difference in the veracity of truth? I seem to remember a quote, "If you tell a lie long enough and loud enough, people will begin to believe it." Does that make the lie a truth? Maybe it makes the lie a fact?

Which "legal responsibility"? When you started this thread, you appeared to be considering how a Court would judge AI, but it now appears that you are talking about writing laws. Although both have "legal responsibilities", they are very different considerations and processes. When creating law, one must consider a large populace, and so, they often look to facts and large sample sizes so the law will be best for most of the populace. In a Court of law, it is the singular truths that are being judged about an individual or circumstance. Very different.

No matter which responsibility you are considering, truth is required to consider it. Why? Because wisdom is required and wisdom is an advanced level of truth. It is not an advanced level of facts -- that would be statistics.

That is about control and how we use AI to regulate the citizens; and therefore, what responsibility AI companies have to the citizenry.

Gee

What makes you think an AI has an opinion?

13 hours ago, Gees said:

Have you had the opportunity to see the movie, "Sully"? It is a true story that addresses this AI expert opinion problem. In the movie, Sully, Captain Sullenberger, is a commercial airline pilot with over 40 years of flying experience, who takes off from the airport and within minutes loses his engines. He manages to land his plane in the Hudson river saving all passengers, but losing the plane. Air traffic control wanted him to turn the plane around and return to the airport, and AI confirmed that that was the best solution, but Sully argued that his experience told him that he would never make it and everyone would have died. Tom Hanks plays Captain Sully and Clint Eastwood directs this true story. I remember reading about it in the paper when it happened 20 or so years ago and remember that Captain Sully almost lost his job, but his experience knew something that AI had not considered, which won his case.

One issue is that “AI” is too vague of a description. That event took place in January of 2009. If an algorithm was involved, it’s not the LLM chatbot that’s being critiqued these days. But I can’t find any mention of this in the summaries of the accident. There is mention of running simulations as part of the investigation.

https://www.ntsb.gov/investigations/AccidentReports/Reports/AAR1003.pdf

Page 2 shows the exchange between pilot and tower. The controller asks if the pilot wants to return to LGA, and the pilot says no. Nothing about an AI. Where did you get that?

15 hours ago, Gees said:

That would be because I am dealing in truths; whereas, Swansont is dealing in facts. Although facts can help one find truth, fact is not truth. Did you ever hear of a Court of law asking for, "The whole fact and nothing but the facts?"

A fact is defined as information or a piece of information presented as true or accurate. To be accepted as a fact, there must be reason and evidence which supports it as a true statement about reality. Facts, then, are very much dependent on their truth value. Your dichotomy is a misrepresenting of what those terms mean. When a Court asks for the whole truth, it is asking for a meaningful and evidenced collection of facts on the matter before the Court.

On 6/6/2026 at 2:59 PM, Gees said:

Have you had the opportunity to see the movie, "Sully"? It is a true story that addresses this AI expert opinion problem. In the movie, Sully, Captain Sullenberger, is a commercial airline pilot with over 40 years of flying experience, who takes off from the airport and within minutes loses his engines. He manages to land his plane in the Hudson river saving all passengers, but losing the plane. Air traffic control wanted him to turn the plane around and return to the airport, and AI confirmed that that was the best solution, but Sully argued that his experience told him that he would never make it and everyone would have died.

I've seen the movie "Sully", and I've also seen the episode of "Air Crash Investigation" (aka "Mayday", etc) about this investigation. I don't recall any mention of AI being used in the investigation. They did use real pilots in a flght simulator to test whether or not it was possible for the pilot for get back to the airport safely. It was initially determined that it was possible for the pilot to get back to the airport safely. However, Captain Sullenberger pointed out that these initial tests failed to take into account realistically the time it takes for the pilot to come to terms with what has happened and to weigh the various options available. When subsequent tests were performed that took into account a realistic amount of delay, none of the pilots succeeded in the flght simulator to return to the airport safely.

5 hours ago, KJW said:

don't recall any mention of AI being used in the investigation.

Yeah, but it “feels” true to Gees and allowed her to support her preconceived narrative. Truthiness is obvs better than facts, and anyone who disagrees is dumb

/caricatured paraphrasing rooted in past experience with this poster

On 6/6/2026 at 5:59 AM, Gees said:

This made me smile. Generally speaking, the purpose of cross examination is to ferret out the lies, but AI does not actually lie, so this seems pretty meaningless. I started to envision a Disney type movie where AI would cross examine AI and it was a lot of fun in my mind, but I am not sure how that could really work in reality.

"To kill a mockingbird" would be a more appropriate type of fiction in this reality... 😉

On 6/6/2026 at 9:08 AM, dimreepr said:

What makes you think an AI has an opinion?

Actually, I don't think it does. I used that word because I was responding to Ken Fabian, who used that word. In law and in philosophy (which are both relevant in this thread), the word, opinion, is often used to name the conclusion that is reached, so Fabian was not really wrong in using it. But you are also correct in questioning the use of the word, opinion, by something that does not have a subjective conclusion. IMO

Gee

On 6/6/2026 at 3:00 PM, swansont said:

One issue is that “AI” is too vague of a description. That event took place in January of 2009. If an algorithm was involved, it’s not the LLM chatbot that’s being critiqued these days.

Agreed -- AI is too vague of a description. I tend to see simulations, AI, and LLM chatboxes, as different levels of the same thing. Simulations are kindergartners, AI is school age, and LLM chatboxes are working toward a degree. Since you are a detail person, I am sure that you can produce reams of information showing how they are different, but don't bother unless you can show me that they are not objective. As that is where I am drawing the line in this thread.

I don't like to see any kind of manufactured conclusions working their way into law and the Courts.

Gee

On 6/6/2026 at 5:59 AM, Gees said:

It is surprising to see how many people do not really differentiate between policy/procedure in business and what is law, and there is a huge difference.

I couldn't agree more. And the bigger the business the more it seems to be able to avoid laws others have to follow.

On 6/6/2026 at 5:59 AM, Gees said:

Although I enjoy philosophy, I rarely post in the Ethics subforum. Thank you for writing an interesting thread.

Glad you like it. I found it difficult to know where to place this thread.

On 6/6/2026 at 5:59 AM, Gees said:

But statistics have everything to do with AI, and how AI affects legal matters is what this thread is about. To my mind, this means that statistics have a great deal to do with this thread and its legal matters.

I don't know how much you know about the British legal system (and of those countries who folowed it) but it is quite different form the American system in several important ways

The British system distinguishing two forms of law breaking, one criminal and one civil.

The burden of proof, penalties and consequences of each are quite different.

In particular the ' balance of probabilities' is not acceptable in criminal cases.

The California case I referred to above could not arise unde British law since criminal responsibility is held by the vehicle owner unless some other person can be proven to have been driving.

An AI is not a legal person, so the responsability immediately falls back on the owner. Statistics has nothing to do with this situation. If the driven vehicle is proven to have run a red light etc someone ( a legal person in British parlance ) must be responsible. that legal person nedd not be Human but cannot at present be an AI.

If only @imatfaal , whose legal knowledge vastly exceeds mine, were still here I would greatly value their opinion on this thread.

On 6/6/2026 at 5:59 AM, Gees said:

On 5/17/2026 at 11:32 AM, studiot said:

Gees has a sample size of 1, but swansont is considering very large sample sizes.

If you tell a lie one time or a thousand times, does that make a difference in the veracity of the lie? Does the number make a difference in the veracity of truth? I seem to remember a quote, "If you tell a lie long enough and loud enough, people will begin to believe it." Does that make the lie a truth? Maybe it makes the lie a fact?

Please allow me the luxury of knowing what I am talking about, mathematically speaking.

Edited Sunday at 10:42 PM1 day by studiot

On 6/6/2026 at 4:23 PM, TheVat said:

A fact is defined as information or a piece of information presented as true or accurate. To be accepted as a fact, there must be reason and evidence which supports it as a true statement about reality. Facts, then, are very much dependent on their truth value. Your dichotomy is a misrepresenting of what those terms mean. When a Court asks for the whole truth, it is asking for a meaningful and evidenced collection of facts on the matter before the Court.

Spoken like a true science guy. Yes, facts are dependent on their truth value, but are truths dependent on facts? No. Instead of arguing this forever, I will give you an example so you can understand my position.

You walk into a room and see me standing over a dead body. A man has been shot and I am standing there with a gun in my hand. Worse yet, the dead body is from a man that I despise and everyone knows it. Six months later, it is a fact that I am a convicted murdered, and it is the truth that I have never killed anyone. I picked up the gun off of the floor because I was stupid with shock, but instinctively I knew that guns are dangerous and there are children about.

Facts require truth in order to be facts, but truths do not require facts in order to be true. This is why science has been called a child of philosophy, but philosophy has never been called a child of science.

Gee

6 minutes ago, Gees said:

Spoken like a true science guy. Yes, facts are dependent on their truth value, but are truths dependent on facts? No. Instead of arguing this forever, I will give you an example so you can understand my position.

You walk into a room and see me standing over a dead body. A man has been shot and I am standing there with a gun in my hand. Worse yet, the dead body is from a man that I despise and everyone knows it. Six months later, it is a fact that I am a convicted murdered, and it is the truth that I have never killed anyone. I picked up the gun off of the floor because I was stupid with shock, but instinctively I knew that guns are dangerous and there are children about.

Facts require truth in order to be facts, but truths do not require facts in order to be true. This is why science has been called a child of philosophy, but philosophy has never been called a child of science.

Gee

In what would would AI be involved in providing evidence in this hypothetical case ?

4 hours ago, Gees said:

Agreed -- AI is too vague of a description. I tend to see simulations, AI, and LLM chatboxes, as different levels of the same thing. Simulations are kindergartners, AI is school age, and LLM chatboxes are working toward a degree. Since you are a detail person, I am sure that you can produce reams of information showing how they are different, but don't bother unless you can show me that they are not objective. As that is where I am drawing the line in this thread.

AFAICT the point of this thread is the issue of some kinds of AI potentially not being objective and the legal implications of that. There are certainly lots of examples of the failures of LLM objectivity, so is that really in question? (Lawyers are being chastised and even sanctioned for using AI to generate documents) I pointed out early on that machine learning and LLMs are not synonymous, and I think it’s a mistake to treat them as such; you need to make these distinctions to have a productive conversation. Is anyone claiming that deterministic or probabilistic simulations are AI?

(and even without that your summary of events of the plane landing doesn’t jibe with the NTSB summary; the simulations were used well after the fact)

As iNow suggested, the objectivity/fidelity/reliability has to be adequately demonstrated before they will be accepted. We have examples of things not accepted in court because they fall short.

19 hours ago, KJW said:

I've seen the movie "Sully", and I've also seen the episode of "Air Crash Investigation" (aka "Mayday", etc) about this investigation. I don't recall any mention of AI being used in the investigation. They did use real pilots in a flght simulator to test whether or not it was possible for the pilot for get back to the airport safely. It was initially determined that it was possible for the pilot to get back to the airport safely. However, Captain Sullenberger pointed out that these initial tests failed to take into account realistically the time it takes for the pilot to come to terms with what has happened and to weigh the various options available. When subsequent tests were performed that took into account a realistic amount of delay, none of the pilots succeeded in the flght simulator to return to the airport safely.

It was a great movie, wasn't it? Captain Sully had worked in investigations before, so he knew the ropes and knew how to get his case heard. He watched the simulations and then noted that there was no time allotted for decision making. The engines failed and the simulation pilot immediately returned to the airport, which is unrealistic. One needs to assess, analyze, and make a decision, which takes some time. The investigators agreed and added 35 seconds to the flight simulator's time for decision, which caused crash after crash. Just 35 seconds between life and death.

It should be noted that at some point, the airlines are going to look at the 35 seconds and the million? billion? dollars worth of airplane. Then they are going to decide that human error is expensive.

In truth the original simulations were invalid because no one is going to turn a plane around before they even know what is wrong and decide to turn the plane around. It made me wonder how many pilots in prior cases were judged to make a "pilot error" because of unrealistic simulations. Anyone can view the Captain's defense in a 20 minute short on YouTube under Sully / A Captain's Defense.

Gee

14 hours ago, iNow said:

Yeah, but it “feels” true to Gees and allowed her to support her preconceived narrative. Truthiness is obvs better than facts, and anyone who disagrees is dumb

/caricatured paraphrasing rooted in past experience with this poster

Truthiness and factiness are not two different teams in competition with each other. Anyone who does not know that is really dumb. Both truths and facts are required in philosophy, but maybe you don't think so. Tell my why.

Gee

12 hours ago, Gees said:

Facts require truth in order to be facts, but truths do not require facts in order to be true. This is why science has been called a child of philosophy, but philosophy has never been called a child of science.

Not at all, facts requires faith for us to accept them, truth is just waiting for us to prove it wrong, again...

7 hours ago, Gees said:

Truthiness and factiness are not two different teams in competition with each other. Anyone who does not know that is really dumb. Both truths and facts are required in philosophy, but maybe you don't think so. Tell my why.

Why are they required?

I can sit here and ponder many thing's and I have no direct information on anything beyond what I think I'm experiencing.

Who taught you that?