AI Hype Machine, Drunk AVs and Space Dogs

Anthony: You’re listening to there auto be a law, the center for auto safety podcast with executive director, Michael Brooks, chief engineer, Fred Perkins, and hosted by me, Anthony seminar for over 50 years, the center for auto safety has worked to make cars safer.

Fred: Howdy gentlemen. Hey, good morning.

Anthony: Morning. Alright. Hey, have you guys ever heard of this thing called artificial intelligence? No, I never did. What is that all about? What do you imagine? Given the name artificial intelligence, it sounds futuristic and sci fi, and it sounds intelligent. What do you think?

Fred: I think it’d be a great idea, particularly since we haven’t found any examples of organic intelligence. Oh boy,

Anthony: the cynic man.

Michael: The problem We’re relying on that organic intelligence to produce the artificial intelligence as well.

Anthony: Yeah, okay one of the problems with artificial intelligence is the name.

The name is stupid, and AI people realized this back in like the late 50s. They’re like, ah, we chose the wrong name, but we’re stuck with it. It’s stupid. They weren’t, because there was only like a dozen people who had any interest in this until about 1980. And then it became two dozen people, and now it’s everybody thinks it’s the most amazing thing, and they think that Computers can think for themselves.

They can think. We’ve all played probably with, Chat GPT and Google’s Gemini and whatnot, and think oh, I typed in something, it gave me a response, it seems intelligent. I want to point out to listeners, there is no intelligence happening at all. It’s really clever programming that basically does statistical analysis to say, hey, here’s probably what would come next in this sentence.

Fair enough to say Fred is our technical wizard. He’s smart.

Fred: I think that’s fair enough to say.

Anthony: Yes, sir. All right. Cross correlation.

Fred: Cross correlation is not the same thing as intuition.

Anthony: Great. I knew my undergraduate degree was not a waste to say as I would come in handy one day. So we have a piece in an opinion piece in the New York Times.

We’re linking to and there’s a It is titled AI and the Silicon Valley Hype Machine. Now, Silicon Valley and hype, come on, that seems a little egregious. But I want to quote from this, there was no pause. This was about basically, researchers saying, Hey, we need to have a six month pause in the development of AI so we can, because there’s a fear they’re going to become too powerful, and will risk loss of our civilization.

Quoting from the article, there was no pause, but now, a year later, the question isn’t really whether AI is too smart and will take over the world, it’s whether AI is too stupid and unreliable to be useful. Consider this week’s announcement from OpenAI’s Chief Executive Sam Altman, who promised he would unveil new stuff.

That feels like magic to me, but it was just rather a routine update that makes chat GPT cheaper and faster. And why are we talking about this on an auto safety podcast? Cause I don’t know. There’s a,

Michael: I’ll tell you, I’ll tell you why. There’s a, there’s another quote later in the article from Molly Walt who Molly White.

Who is a cryptocurrency researcher and critic and she’s, she, this quote is great to me because it brings up a lot of familiar feelings. She says, I find my feelings about AI are actually pretty similar to my feelings about blockchains. They do a poor job of much of what people try to do with them.

They can’t do the things their creators claim they one day might. And many of the things they are well suited to do may not be altogether that beneficial. Y’all. Which to me sounds almost exactly like what we’ve seen going on in the autonomous world and in the Tesla world of pretend autonomy.

Fred: Let’s remember that these pronouncements are generally coming from the CEOs of these companies and the CEO’s responsibility is to pump the stock price.

That’s what they get paid to do and every word that comes out of their mouth. towards pumping the stock price. So the relationship to reality is often pretty tenuous.

Anthony: So I, so you’re being cynical again. You think that these shareholders and investors and institutional analysts are gullible.

Michael: I think that they’re not concerned about safety. They’re concerned about money.

Anthony: This is a great opinion piece and it’s going to tie into the next thing we’re going to cover is from MSN. We have a link. Titled, Driverless Waymo Vehicle Glitches While Driving in Metro Phoenix. Watch the video.

And these cars are using some level of artificial intelligence. Again, more artificial than intelligent. Very clever. Mathematical inference machines, essentially, that say, Hey, I’m gonna guess this is what should happen next, right? So in the, according from the article, In the past, Waymos have been spotted stalled in downtown Phoenix and routinely idling in certain neighborhoods.

However, a new video getting attention on social media shows a Waymo repeatedly swerving to the bike lane while driving in Phoenix. This video is Awesome. And it it’s

Michael: not, no, it’s not awesome. It is a big yikes.

Anthony: It’s awesome because it goes against the argument that people have for self driving cars all the time.

And that number one thing is autonomous vehicles don’t drive drunk. But if you watch this video, the car is drunk.

Michael: Yeah, the Waymo’s drunk.

Anthony: Yeah it’s fascinating. And so this is part of Waymo being investigated. Sorry, Michael.

Michael: I was just going to say, this is, it also hard to tell exactly what’s going on there.

If it’s struggling to see around the trailer in front of it or what, but it is pretty clearly not identifying the bus and bike lanes, it looks like it’s a, one of those wide lanes at first where buses and bikes travel, and then it converts to a bike lane only. And the Tesla, the Waymo is not only crossing the.

Into the bike lane, it occasionally crosses into, over the divider in the center lane. So that’s a significant problem. And yes, that, that vehicle looks like it’s being operated by someone who is impaired.

Fred: Is that really a problem? Because, I would think that before they put these vehicles on the road, there’d be some way for law enforcement to stop the vehicle if it’s acting erratically.

Isn’t that part of what they have to do, Michael? Isn’t that a requirement?

Michael: I don’t know what that would charge the vehicle with if they pulled it over. The solution here, would be to box that vehicle in and not allow it to move until you’d contacted Waymo. And they came and picked it up, take it off the road, but there’s, it raises the question, if you put every Waymo vehicle on that stretch of road, are they all going to be doing the same thing?

And what’s the problem here? It’s pretty. It’s a pretty big gap there between what you hear the A. V. Industry saying about how well they’re performing and what we’re seeing on the road. This is just the latest example of that.

Fred: So your answer is no, there’s no way for law enforcement to stop it.

Is that correct?

Michael: There’s no way for law enforcement to directly send a signal and stop the vehicle. I assume they could, call Waymo. I believe Waymo was the company that recently said they are, putting The ability for first responders and people to actually talk to the vehicle to communicate with the road, remote operator when those vehicles are having problems.

So, they might be able to eventually stop that vehicle. But it’s. I don’t know, there’s not a button that they can hit and stop it immediately for driving like that. Similar to how, the same thing exists when they’re trying to stop a human who’s driving drunk, that human might pull over.

A human might not pull over. And sometimes you might have to resort to some interesting tactics, maybe not so safe to get that driver to stop.

Fred: And maybe we could approach HBO and, do a series on drunk AV history.

Anthony: Oh, I love it. I think this is a good one. But this brings up a question, so I hear regularly outside my window, the police get on their, I don’t know, their loud hailer, whatnot, through their car and be like, Pull over.

Go past the traffic light. Pull over to the right. Pull over to the right. Pull over to the right! So they’re regularly doing that. It’s I have PTSD from it. So if a cop is behind one of these vehicles and they get on, their megaphone, and they’re like, hey, pull over. Pull over.

Pull over. There’s no one in the car driving it. There’s a passenger in the back playing, Wordle. And I imagine with Waymo, they have remote operators, but they don’t have, it’s not a one to one relationship. So someone gets an alert, maybe on a screen, and is Oh, wait, what?

Michael: Yeah I don’t know how that works.

You would have to assume, I know that the police in Phoenix probably have a phone number and all of their phones that tells them, call, here’s the number for Waymo, one of their vehicles acting up, call this number and the remote operator can then, either take over and stop that vehicle or, Take it out of service or whatever they do, but it’s something that all the autonomous vehicle companies really need to work on is communication with local law enforcement, fire departments, emergency services.

We’ve seen this before on a grand scale in San Francisco and all the problems they had there. And Waymo has. It’s performed better than cruise in a lot of metrics, but, we’re definitely starting to see some issues arise this, the NITSA opened the investigation, I believe it was last week when, into that incident and many other incidents, but I think particularly in the incident when the Waymo was crossing the center line to pass the unicyclists and now we see another Waymo doing something that’s every bit as dangerous.

And it’s a big concern and it’s, it becomes even bigger when you realize that the industry groups that are supporting this technology are out there pretending that none of this happens.

Fred: Oh, Michael, it sounds like you’re describing Godzilla running rampant through the street,

Michael: beyond the

Fred: control of security forces.

Michael: With the rainbows are not breathing fire yet, and they’re not stepping on buildings, but I beg to

Fred: differ. They do catch on fire.

Anthony: I think all of these AV vehicles need to have what are on a lot of 18 wheelers and the back says, how’s my driving? Call this number. You’re driving shit.

It’s scary. That’s how you’re driving is right now. Right now, you’re driving drunk and you claim you don’t. It’s almost like we’re being gas lit. Ho. Oh ho.

Alright next story in Ars Technica. Tesla must face fraud suit for claiming its cars could fully drive themselves. Speaking of being gas lit, Elon Musk, everybody. The world’s greatest gas lighter. BS artist. A federal judge ruled yesterday that Tesla must face a lawsuit alleging that it committed fraud by misrepresenting the self driving capabilities of its vehicles.

Basically, for those of you playing the home game, since 2015, Elon Musk has said that self driving is a solved problem, it’ll be ready next year. 2016, it’ll be ready next year. 2017, it’ll be ready next year. 2018, why are people still listening to this guy? Further in the article, the complaint also sufficiently alleges that mess Musk, falsely represented the vehicle’s future ability to self drive cross country, and that Lo Savio, the plaintiff, relied upon these representations pre purchase.

The judge concluded I think, I got confused there, sorry. Musk claimed at an October 2016 news conference that a Tesla car would be able to drive from Los Angeles to New York City by the end of next year without the need for a single touch. I think somebody who doesn’t need a single touch is Elon Musk.

Yeah fraud? Not fraud.

Michael: I think it’s clearly fraud and we were calling it fraud, right around 2016, 2017, what was happening. It’s pretty clear the Tesla website was saying, these vehicles come equipped with the hardware necessary for fueling. Full self driving and it’s 2024 and we still don’t think that the vehicles they’re selling come equipped with that.

And they certainly didn’t in 2016. And I, it’s a massive fraud. I don’t think there’s any other way to represent the situation. Tesla has made. A lot of money and Tesla stockholders have made a lot of money off of the investments made in the company based on this fraud and you know now you see musk even doubling down on the robo taxi joke that we just don’t see happening and I don’t know what else to call at this point other than a massive fraud.

And everyone who paid for full self driving you’re still getting a car. So the car itself is not fraudulent. You’re getting a car, but if you paid for the full self driving you should certainly be refunded that amount because you’re never going to get there. And then the vehicles they’ve produced so far.

Anthony: Are you a listener? Are you a Tesla stockholder, shareholder? Have you made over a thousand percent return? Because if so, you can donate to the Center for Auto Safety with some of your winnings.

Fred: Oh, good segue, good segue.

Anthony: I hope you sold before the recent 60 percent drop in. But seriously, donate.

That’s good. Another Ars Technica article compares Waymo and Tesla. And it’s titled, On self driving, Waymo is playing chess while Tesla is playing checkers. I think they’re both playing a different game. It’s called, Screw you the public! That’s pretty much it. I think

Fred: you would be facing a firing squad versus Russian roulette, I think.

Oh, excellent. Michael, what was your take?

Michael: Yeah, I could see that. That makes sense too, because, I think my take is that if you’re, if I had to choose a vehicle to hop in and ride around town, and right now it would be a Waymo a thousand times over a Tesla with full self driving. Simply because.

I know that Waymo has better sensors, appropriate sensors, maybe, for the circumstances. Waymo probably has redundancies in place that I don’t believe Tesla’s willing to invest in. And, frankly, I’m going to say the T word, trust, here, that the AV industry keeps repeating. I trust Waymo more than I trust Tesla.

Anthony: Fair enough. And so the take on this is that Waymo, their approach is saying, Hey, we’re in this, operating these limited operating design domains. We have redundant sensors, we’re not just relying on cameras. And we have remote operators that’ll get involved. And right now, this has been more or less okay.

Because they have, a lot more remote operators involved who can take over and intercede because they’re happening in metro areas with good cell phone coverage and what not. But, when they’re traveling on a freeway at 60, 70 miles per hour, somebody sitting in, Kalamazoo is not gonna be able to respond in time.

You’re dead. I’m sorry, you’ve hit the br barrier, you’ve hit whatever. It, there’s just Physics, there’s no way for that remote operator who’s probably watching dozens of vehicles to be like, Oh, let me get, Oh no, Oh, I missed that one. It’s like the worst air traffic controller.

Michael: And there’s some there, there were some interesting things in this article and then further evidence of a difference between Waymo and Tesla, Waymo programs its vehicles.

When they encounter a problem, they, basically, if, and they don’t think it’s safe to proceed, they basically slow a stop and call into headquarters and say, something’s wrong here. Whereas a Tesla tries to shift the burden over onto the driver who should be monitoring the system, but isn’t always doing a good job of that.

And it was interesting, that Waymo is saying that remote operators don’t directly drive those vehicles. Essentially, when a Waymo encounters a problem and pulls over to a stop and calls into headquarters, it’s not calling in for. An operator that’s going to pick out, take up a joystick and control the vehicle and get it through the situation.

They basically operators, the remote operator answers questions that the car is asking and builds the vehicle’s confidence that it can move in the right direction, which is an interesting way to do it. But it seems like that might get around some of the problems we’ve identified with latency because you don’t have an operator directly controlling the vehicle and that problem with a delay in commands or delay in detecting objects or other.

Problems that are in the vehicle’s path,

Fred: Michael, you’re exactly right. How unusual is that? This is the same. This is the same technique basically that NASA is using to control very remote operations in space. For example, it takes several minutes for a signal radio signal to reach from Earth to Mars.

What they do is they make sure that the reentry vehicle has got the right trajectory. They make last minute adjustments, but then the vehicle is on its own. It’s got a, it’s got its own control system. It’s got to react instantly to whatever circumstances it finds. This is actually an excellent way. To set up a control system only time will tell whether or not waymo has enough capability on board to handle the complexity of open traffic and unregulated, uh, traveling through all kinds of streets and circumstances, but it’s actually a very good way to set up the control system.

And hats off to them for doing that. It’s very difficult as well. It takes a lot of engineering, but it is Potentially an excellent solution.

Anthony: But is this a good approach when on freeways when you’re, you’re traveling at 70 miles per hour and some tractor trailer jackknifes in front of you?

Michael: Yeah it’s hard to say there because the vehicle’s it’s not, the vehicle’s not really going to stop, you don’t really have areas to stop and wait for commands on interstates. You do have a shoulder on some interstates, not all in the cities. It’s less likely that you’re going to have them where Waymo is operating.

And the speeds involved. You simply, if you encounter an issue. You’re probably not going to be able to dial home and get her to get an answer before you’ve encountered that problem. So

Fred: no, I think that’s yeah, that’s completely right. There’s no way you can expect remote driving to handle instant hazardous conditions on a highway, right?

Anthony: Because I’m imagining, some space vehicle coming back and re entry. There’s no options of somebody walking their dog in front of it. There’s no, everything in the sky is going to be known.

Fred: Actually, no, I’m just but not really, because you’ve got weather, don’t forget, and weather is your winds and crosswinds.

And who knows what you’re going to encounter on the way down. No, there’s no space dogs up there, but there are variables. And also, when you are about to land, you don’t know whether you’re going to land in a gully or on top of a mountain. So there’s decisions that need to be made in the instance before touching down.

So yeah, no space dogs, but still a lot of variables.

Anthony: No space dogs. Yeah, I’m really upset now. Wait, no, the Russians had a space dog. Come on. Oh, poor Laika. This article goes on talking about how the Tesla fan folks are saying that, hey Tesla’s approach is better because they’ve got nonsense coming out of their mouth.

And on a favorite neo Nazi holiday coming up in August, Elon will release the robo taxi. I think it’s fair to say if. You’re listening to this, you you’re probably sober right now, and you’re probably not delusional, and you realize that the robo taxi won’t happen. At best, they’ll have a guy dressed up in a robo taxi costume who’ll dance for you.

Fred: But listeners, if there is somebody who’s got a drinking game associated with this podcast, please let us know. We’d love to know how that works.

Anthony: All right. That’s enough of our AVs for this episode, I believe. Let’s jump on into the law. The law. So we’ve talked about this a number of times in the show about, hey, cars are now computers on wheels, and they’ve got cameras and microphones, and they’re listening to you just right now.

You’re listening to us. Oh my god. Now it’s coming out Automotive News recently reported that eight automakers sent vehicle location data to police without a court order or warrant. The eight companies told senators that they provided police with data when subpoenaed, getting a rise from several officials.

BMW, Kia, Mazda, Mercedes Benz, Nissan, Subaru, Toyota, and Volkswagen presented their responses to lawmakers. Senators Ron Wyden from Oregon and Ed Markey from Massachusetts penned a letter to the FTC urging investigative action. Automakers have not only kept consumers in the dark regarding their actual practices, but multiple companies misled consumers for over a decade by failing to honor the industry’s own voluntary privacy principles.

This is bad. We don’t want your car to spy on you.

Michael: No, it’s, it points out the problem and it’s something the auto industry does across its operations, not just in, in privacy, but the industry is always proposing voluntary. Guidelines and principles. They did it with automatic emergency braking.

They’re currently doing it with cybersecurity. They want to make their own laws. They don’t want government regulators telling them how they’re going to operate their businesses. So they often try to preempt government action by saying, Whoa, Whoa, wait, we’ve got this taken care of. Just trust me.

us, we’re going to do this, and it forced us against, the members of our coalition, and you know what, here you go, 10 years ago, they basically said, here you go, FTC, here is, here are our principles, if you see us breaking these principles, feel free to pile on to us and do something about it. It took the SEC about 10 years to do that.

And they’re slowly, the FTC slowly seems to be coming out on this issue and trying to enforce some sort of, of consumer privacy protection for people who are buying vehicles and aren’t told what those vehicles are collecting and how they’re distributing that data.

Anthony: So I have a car, I’m driving around, it’s collecting not only where I drove, how fast I drove, breaking events.

We’ve discussed, I can’t remember which manufacturer. Is collecting your sexual preferences a whole bunch of crazy stuff. Isn’t there some sort of protection for me as a consumer? It’s my car, I own it. Don’t I own the data that my car is collecting?

Fred: Oh you’re so cute, Anthony.

No, it’s wild west.

Anthony: How is that I’m looking anything else I own you know, isn’t it my data? Isn’t it my information? And isn’t it, wouldn’t it be illegal to just hand this over to anybody who asks for it?

Fred: Apparently not. You don’t own the software in your phone either. No, it’s licensed.

Michael: I think, and the problem here really, you probably don’t own your data. You may own a part of it and but when you buy that vehicle, you’ve often signed or initialed or checkboxed away. Your sole right to determine and control, what happens with that information. And a lot of, there’s some interesting parts of the report that it was Senator Ron Wyden, who led this effort, and they put out a report that goes through what happens with each automaker in certain situations.

And. There are, this kind of data can be used in law enforcement investigations, for instance. And, the say the FBI is investigating a drug trafficking ring, and they want the locations of certain individual so that they can Figure out if that person was in proximity to a known delivery spot or whatever it might be.

I believe it was, there’s only one manufacturer, that’s Tesla, which tells its customers when law enforcement or a judicial subpoena or government subpoena has been issued seeking information on that person. All of the other manufacturers don’t provide their customers with any such notice. So that’s.

I guess that’s a point in Tesla’s favor. Although essentially the consumers just don’t really have a good idea of what the information is that their car is collecting and what the companies are doing with it. And, that’s a two way street here, folks. If you’re buying a car, you need to do your due diligence here because manufacturers are perfectly willing to take consumer A naivety.

How do you pronounce that word? I don’t know, but they’re willing to use that against you so that they can collect your information without you knowing it and monetize it and do other things with it. So it’s critical for consumers when they’re buying a vehicle and particularly connected vehicles these days.

To be in to, to understand going in that there is going to be data collected on you and that you need to opt out where you can, if you don’t want that data being collected, and you need to understand that there may be some things that are being collected that you’re not going to have control over, that, that may be subject to subpoena may be subject to all sorts of.

Leaks. I’ve experienced at least two data leaks in the last year that exposed all of my information to the dark web. And that is a huge possibility for, your vehicle GPS data or for your contact list or anything that’s stored in the vehicle. And so it’s critically important that consumers be aware here and not trust that this information is going to be protected.

Fred: We had a story last week about a Tesla Cybertruck that had been stolen, and the owner was able to trace it very rapidly by tracking its data and vectored the police into the Cybertruck, which was then recovered, and the thief was apparently arrested. But it’s important to note that If the, if a bad actor had equal access to the data in the cyber truck and someone else’s cyber truck was able to, track and understand exactly where the occupant was and what they were doing, that could be a really bad thing.

So the capability of tracking your truck when it’s stolen has got to be modulated so that bad actors cannot hijack the truck, cannot hijack that data capability, and do bad things to your property or yourself.

Anthony: I wonder if there’s a new app that’s going to come out, basically like Adblock for cars.

Because in my home network, I can, My internet provider and everybody else will try to get the data. My, my TV has like an ethernet port in it. And I’m like, I’m not hooking that up. Why do they need to know? There’s gotta be a way to do that with cars. But, and Michael, I apologize again for selling your information on the dark web, but you know.

Not enough people went to autosafety. org and clicked donate. Ha, you do that, I’ll steal that

Fred: from Michael.

Anthony: In a related New York Times article about this, a lot of people ask who cares if my data is collected out there, I’m not doing anything wrong. Problem with that is, what’s the definition of wrong?

Or how people are doing this. So there’s a great quote from this New York Times article. As far as wrong, As far right politicians escalate their war on women, I’m especially concerned about cars revealing people who cross state lines to obtain an abortion, Senator Wyden said in a statement. And this is a perfect example of how your data can be used without your consent for very private reasons and whatnot, and that’s that’s frightening.

There needs to be, there ought to be laws for that.

Michael: There ought to be a law for that. It’s it’s very concerning, that you could have, a state can, at this point can issue a subpoena or, demand essentially the data on someone who leaves that state for the purposes of acquiring an abortion, which is a, it’s not really our subject here at the Center for Autosafety, when Privacy.

Data, data is being used that way. It’s very concerning that a state actor can take that kind of path. And I don’t know, it’s, it feels, it just feels creepy to me that we’re entering an era where that’s possible.

Anthony: Yeah speaking of more creepy, let’s go to a different angle. I think this is sad.

It’s not something we ever talk about. But we all know who made our phones and who made our clothes. And it’s something we just pretend isn’t true. Automakers got caught with this. A congressional investigation found that BMW, Jaguar Land Rover, and Volkswagen purchased parts that originated from a Chinese supplier Flagged by the United States for participating in forced labor programs in Xinjiang, a far western region of China where the local population is subject to mass surveillance and detention.

Ah, Basically, they’re using slave labor to make your cars. Isn’t it bad enough that they’re making our clothes and our phones. Now they got to make our car. This is too dark for this time of day.

Michael: No, this is something we discussed, a while back we talked about the Uyghurs and the forced labor in China and there, there is a law that was passed, I believe in 2021 to address this.

And Ron Wyden and a number of senators looked into this issue and found that, automakers are basically sticking their head in the sand and they swear they can’t find any forced labor in their supply chains. And one of the issues here is that supply chains are massive. Every vehicle has.

30 or 40, 000 parts. And you’re having to build those parts all over the world. And yes, it is a massive undertaking to keep track of your parts. But the fact is the automakers are the only ones who can do that. And they’re the only ones who can verify that they’re not getting parts from places where forced labor is taking place.

And so it’s their responsibility and no one else’s. To make this happen. And, they found that BMW had actually imported cars containing parts from that’s area in China where there is a prohibition on, on, on manufacturers using parts from that area Jaguar Land Rover. And they imported parts from that area and Volkswagen manufactured cars with those components that were.

From a supplier band for Uyghur support Uyghur forced labor. So another area where the automakers aren’t doing a great job policing themselves and essentially, if they’re not policing themselves in this area, then they are. Supporting forced labor in China. And it’s something that they, I think, and I think I agree with Mr.

Wyden that they need to do more here and that they’re trying to ignore a problem that’s critical.

Fred: Well, egregious labor practices are not limited to China. We remember last year, I think it was last year that it was reported that Kia and Hyundai were using illegal child labor. In the factories and where’s the Mississippi and Alabama?

Michael: It was Georgia and Alabama. I believe it was have a plant Mississippi. Yeah. And it was when they were essentially They were, I believe it was, they suppliers surrounding the Kia or the Hyundai plant. I can’t remember which one it was. I think it was the one in Alabama and Montgomery were employing underage children.

Some of which were, some of whom were even involved in, software programming and other things. That sounds like a really shady practice. So yes, the only reason that, there’s a law that prohibits manufacturers from using components from the areas in China where forced labor is happening, but there isn’t a law preventing the import or the manufacturer of vehicles from anywhere else on earth where those practices might be taking place.

So this is, that area of China was a particularly bad area. For that, automakers are probably and most likely, I’d have to say they are using forced labor in some form in a number of countries in the world that may not be China. And it’s probably an area where there needs to be a lot more law in place, and we really need to get automakers to do a better job in their supply chain of ensuring that they’re not, supporting slavery in many respects.

Anthony: And I think as a consumer, the best thing you can do is continue to listen to this, share it with your friends. Because we’ve rattled off a lot of manufacturers names today, but I’ve noticed absent from the list are the three U. S. car companies. Ford, GM, and Chrysler Stellantis. Is that what they’re called now?

But they, they don’t seem to be the one. They’re in France

Michael: now, essentially. Although they, they manufacture a significant portion of their vehicles and parts in the United States. They are a company.

Anthony: All right, let’s they were Italian

Michael: and they were German before that. So Chrysler’s had a long road.

I was just gonna

Fred: say, there’s a big now, .

Anthony: He and now that you, Fred, you’re talking, let’s go to the Tao of Fred and a new feature we’re gonna talk about. That’s separate from the ta, but Fred’s gonna introduce it in this week’s ta. And is that confusing enough for me and the AI software transcribing this?

That’ll do it as the tower. You’ve now entered

Fred: the Tao of Fred. So the new section is going to be. Gaslight Illumination, because there’s so much gaslighting around, but it simply doesn’t get the publicity that it deserves. And so what we’re going to, it’s going to be a reader participant or a listener participation game, by the way.

So we’ll be looking for your input as well. But to give you a sample of the kind of thing that we’re going to talk about. We are fortunate to have listened to lies and misrepresentation by Jeff Farah, Chief Executive Officer of Autonomous Vehicle Industry Association, in Senate testimony yesterday.

He was, sorry? I said, oh boy. Yeah, he’s it’s interesting how people just are unintimidated by the need to present accurate information to Congress. I guess that’s a new thing. Anyhow, a couple of instances, he says that our mission is to bring the tremendous safety, mobility, transportation and V’s. To consumers and businesses in a safe, responsible, and expeditious manner.

That’s great, but there is no evidence that there is any tremendous safety, mobility, transportation, or economic benefits of a DS. Accept the ability of extracting dollars from investors in these companies that actually are scammy. It goes on to say this is all on the 1st page. The vehicles are operated by a, b, I, a members autonomous vehicle industry association members have driven nearly 70M miles on U.

S. public roads. That’s actually interesting because it was 7 million last week, and then all of a sudden it became 70 million, but there’s been no outside review of where those miles occurred, where the unrestricted roads with unrestricted passes that open highways. We don’t know where it was, and we don’t know what the circumstances were, but what we do know.

Is that 1 of the AVI members is a company called Uber. The other many others, but they include Waymo, Cruz, Aurora, and we know a lot of these companies have problems, but we know that a pedestrian was killed by Uber while operating on autonomously on U. S. public roads. For them to say that. They have demonstrated the tremendous safety is absolutely unsupportable because even if you look within their members, there has been a death and serious injury by cruise as well as a lot of other accidents within that unsupportable 70, 000, 000 miles of that they’ve driven now for reference.

Conventionally driven vehicles have a fatal crash roughly once every 110, 000, 000 miles. We’re looking at, even if you give them the benefit of the doubt that those 70, 000, 000 miles are representative, you’re still looking at least. Roughly twice the fatality figures for these in the members as for conventionally driven vehicles.

These claims of safety have absolutely no support. But the testimony goes on and continued gaslighting. It’s actually a very interesting example of gaslighting because so many of the A. V. Documents that are around there’s so many disquisitions or speeches, whatever that are around, they very carefully mix tenses.

And it’s very misleading because they say that, for example, 80s are poised to significantly improve roadway safety as they do not speed. They do not text and they do not drive while impaired by alcohol, drugs or fatigue. That’s fine, but they’re also slow to respond to complex stimuli and they act as though they are drunks.

They often misinterpret their circumstances, like drunks. They swerve uncontrollably. We talked to one of those videos this morning. They make wrong turns like drunks. We have videos of way modes going up the wrong side of limited access highways. They are easily distracted, particularly by things like photos on the side of trucks and text on the side of trucks.

And any 2 year old or not, maybe not any 2 year old, but typical 2 year olds can look out a window and say, there’s a truck. There’s a fire truck. There’s a school bus. They often fall in love with these things. AVs have a lot of difficulty doing that because AVs see a scatter in your data. They’re easily distracted by discontinuities in that data, and it’s a big challenge for them to interpret, for example, the side of a truck that has a picture of a plumber and text that says stop leaks from a person crossing a road next to a stop sign the trucks or the AVs struggle with that.

So this whole idea that they’re never distracted. It’s unsupportable, it’s completely wrong. They then go on to talk about a canard, or I should say he goes on, Mr. Farrar goes on to say, by removing human error as a cause of roadway accidents, AVs can help reduce roadway deaths, saving the lives of countless Americans.

That goes back to a canard of a misinterpretation of a study that NHTSA did on crash causality a long time ago that said 94 percent of the crashes that occur. Have a critical component that involves human beings. The correct way to interpret that is, and I thank Phil Coopman for this because he discussed this week that if human beings had known what was going on and hadn’t able to respond in 94 percent of the circumstances, they might have avoided the crash.

That’s a long way from saying the humans caused the crash, right? It’s what it’s really saying is that things happen so fast. And unfortunately, that the humans couldn’t respond in time. Very different take on those numbers. It’s very important to demythologize this whole idea that the AVs are going to reduce highway deaths by 90 some odd percent.

But it’s also important to note that by removing human judgment from driving control and exposing the public to a wide range of lethal data processing and novel personal security issues, for example, authorization of commands, identification of the person giving the command and cyber security, the hazards of death are greatly increased.

In ways that are simply not present with a human driver, but, again, going back to whether or not any of this is supportable the statistics on death and injury caused by a members is greater than much greater than. The numbers that are out there for human drivers, even including the really bad human drivers that are out there.

There’s just nothing here that’s supportable, but then they go on to say things like access to food is another area of equality that AVs can help alleviate. They cannot help alleviate that. Okay. It’s simply not possible. Now, maybe someday in the future pie in the sky, by and by, they would be able to do that, but they cannot do that.

Anthony: I’m sorry, but by access to food, is this the the self driving pizza truck? Oh, man.

Fred: Yeah. Nuro does that, right? They deliver pizza. But I think there’s only a couple of square miles where they do that. But, they just go on to talk about an endless list of benefits as though they are actually out there and doing that.

AVs can prove particularly useful for improving access to food, blah, blah, blah. They cannot do that. They don’t do that. Maybe someday finally, they go on and say something. The integration of AVs into America’s commercial fleets will help optimize the transportation of freight nationwide. No support for that.

Absolutely no evidence that has ever happened and will it ever happen in the future? Could be but at the moment, it’s all bullshit. It’s All gaslighting, and it says for consumers, the are positioned to reduce general transportation costs. Where is that? These costs hundreds of thousands of dollars to operate, build and operate and maybe there will someday be a robo taxi, but the economics of such robo taxis and the economics of reliance on a transportation.

Are simply not supported by any business model that’s based upon the actual capabilities of these vehicles. So when they say use the present tense are positioned well, what they really should say babies might buy and buy, pie in the sky do some good for these things. It’s bad pie in the sky by and by.

It’s important to discriminate the Pittsburgh from the reality. Now, finally, they do go on to say something that’s honest. And towards the end, they say the American Navy industry is that an inflection point as the technology is not being commercialized and the benefits of are beginning to accrue. Maybe at least the tenses is correct are beginning to improve, but whether the benefits are beginning to accrue or the actual penalties associated with the use is still unresolved.

So I leave you with that. That Mr. Farah. Gosh, it just has no shame and good for him. I guess that’s what the AV industry is all about right now.

Anthony: Big shout out to Jeff Farah, this week’s gas pass winner. Gasiest, most flatulent individual of the week, something like that. We’ll work on that. So one thing you briefly touched on there was the autonomous trucks and the 18 wheelers.

And this is a question I have, cause I believe it was Volvo is coming out with their production ready, self driving 18 wheeler truck. Aurora

Fred: Volvo. Yeah.

Anthony: The Aurora thing. And we’ve talked about Aurora in the past and like their, They’ve got all the sensors and the whole thing, and they’ve been running tests, and hey, this is gonna save money, cause we won’t need a driver anymore, and it’ll go from hub to hub.

And it got me thinking that wouldn’t it be cool if we had something, like, where you can set up having an 18 wheeler that you could pull I don’t know 50 cargo containers on dedicated path that you’re not going to interfere with traffic and go from hub to hub. And maybe you just have all those 50 things, you just got two people monitoring it.

And we could call it railroads. All on a train.

Fred: What a great idea.

Anthony: It’s

Michael: a great idea, unless you’re a Tesla on full self drive.

Anthony: That there’s not going to make it why, this is very much off topic, I think for an auto safety discussion, but, uh, why aren’t we just using rail for these things?

Get the 18 wheels off the road. We have a rail network. Just do that. Okay. We just broke, we broke Michael. So earlier today, Brad hit him with something, he stumbled. Then Brad hit him again with something else and he stumbled and his face just compressed in the most interesting way.

Michael: Going into the, I don’t know I liked the idea of trains.

Running everything we need them to run across the America and not having to deal with 18 wheelers as I drive down I 81. But the train industry has its own issues and it, it’s, it, I don’t know. And maybe that’s a solution. It’s, that’s a tough question. That’s why my face creakled up.

Cause I don’t have any answers there.

Fred: Yeah. But let’s remember that. When a motorist is on the road or pedestrian is on the road if they’re in the proximity of a train. They generally know it, the train stops the horn before the crossing gates go down. They make a lot of noise. Yes, there are still occasionally collisions and crashes between people and trains sometimes intentional, but a lot of, a lot of energy and thought has gone into protecting people and vehicles from the hazard associated with the train.

There is no thought. going into protecting people from the hazards associated with self driving trucks, except to say, trust me, we have a great control system and gosh, we wouldn’t do anything because it would harm our business model. There’s just there’s no there.

Anthony: There’s no there.

Let’s move on to something, one more article before we get into recalls. This is always a complaint about electric vehicles. That, oh, the battery’s gonna die and then you gotta replace the whole battery and that’s 20 grand. These things are dumb. And I’m all for electric vehicles, because in my mind, I’m gonna live with a place with a full solar farm, or, my energy’s gonna be, powered by dolphin smiles, something really good, and there won’t be any problems.

And, So here’s an article from K U O W and it’s someone talking about their Tesla Model S from 2013. How they’d driven up from Portland to Seattle to visit their daughter, and when it was time to head home, she turned on her car and was met with a blank screen. Tesla ran diagnostics and discovered the battery was fried.

They said it needs to be replaced and that’s expensive. Expensive meant as much as 20, 000, which is basically what the car’s blue book value is worth at the time. And now that sounds crazy, but I think Tesla wants to sell you a new car. And this article talks about a little repair shop where what they do is, No, we fix these batteries, and what this does is there’s not enough people trained on how to fix these batteries, So instead of replacing the entire thing, They go in there and they find the cells that have died and broken and just replace those little parts. Which makes a lot, 20 grand. Which makes a lot more sense. And this is something that we’ve talked about before, whereas there’s not enough independent repair shops to that have been trained on this, or even manufacturers allowing people to be trained on this, to open up these battery packs to replace and repair.

Michael: Why would they, if they can just sell someone a new vehicle, right? They don’t have an incentive to support a network of highly skilled technicians who can repair batteries when they can sell that consumer another battery that’s going to cost them 20, 000. It’s.

Anthony: What’s happening there?

What was that sound? Oh my God.

Michael: But there, we, we also have questions about whether, having folks repair damaged batteries could present, safety issues as well, because that battery is no longer been, at least. Pass the manufacturer’s safety checks before, before it leaves the manufacturer.

So this it’s a very new industry repairing electric batteries. And I, we want it to get much better, much quicker, but also we want the batteries to get a lot better so that we don’t have those safety issues we have to worry about. And right now we’re not there.

Fred: Remember the job of the CEO of these companies and.

Musk, in particular, is to pump the stock price. There’s not much about repairing these cars, making them repairable, that helps the stock price. In fact, one of the important tasks for Musk right now is to make sure that nobody notices that the resale value of Teslas often drops rapidly to zero. If you think of this one as an example, all right, the Kelly Blue Book value was 20, 000.

Cost to replace the batteries, 20, 000. 20, 000. Michael, you’re a math genius. What’s the net value of the car? It’s zero. And this is why this is becoming a widespread phenomenon in the EV industry. And if you look at the resale value of the EVs it’s not strong. People are aware that they’ve got a big financial burden associated with the aging of these vehicles.

Michael: I think there’s a, there is one large participant in the vehicle area that wouldn’t be a fit here, which I think insurance companies who are having to essentially pay for new batteries in every time a vehicle crashes, batteries are being replaced, even in some, somewhat minor collisions.

And that is really hitting insurers and car insurers hard. Electric vehicles have significantly increased their payouts. And we, I think we discussed a couple of months ago with how the insurance industry is no longer even making money surely on the amount of the policies they’re putting out there having to do so through investment.

And so it’s a problem. It’s a significant problem for the insurance industry and they might, be willing to support a training and expanded battery repair programs across America that could prevent consumers from running into the situation of having a 20, 000 battery replacement.

Maybe they have, a couple of thousand dollars that they can invest in a battery repair and drive the car for another five years.

Fred: I like that better. But there are companies that are Orienting themselves around replaceable batteries routinely rather than having cars to recharge themselves so they could supply our batteries that they can swap in and out.

That would be a much, much better model than the current technology, which basically seals the entire case and requires you to do a lot of work simply to open the case and access the cells. They’re not designed for repairs, and we talked about this for Rivian, how it was not designed for repairs, and they’re getting extraordinary astronomical repair figures for relatively minor bumps and bruises as the vehicles suffer.

As part of the maturing industry, which is slow to recognize the responsibility to allow people who deign to purchase their vehicles to repair them economically. Alright,

Anthony: well, listeners, check out this article on KUOW. The sound of the valley. I have no idea what that station is. How about some recalls?

Big smiles, heads nodding all around, cheering, it’s amazing. Fred’s lighting balloons on fire. I don’t know why. BMW, potentially 3, 256 vehicles. The 2025 BMW X5 S Drive 40i, X Drive S Drive there’s too many. You gotta be, I’m not gonna list them. They’re an S, BMW, they’re an SUV. They may not can fully conform to, Federal Motor Vehicle Safety Standards, 208, Section 2.

3, Section 14. 5. 2, and Section 15. 2. These are some of Michael’s favorite and least favorite parts of the Federal Motor Vehicle Safety Standard. Here

Michael: they are not compliant with FMVSS 208. That’s basically controls airbags, seatbelts, any occupant protections like that. And what’s happening here is the sensor thinks that An occupant is belted when they’re not, and so they don’t get the annoying reminder.

And, they might end up in a crash unbelted because of that. That’s something we’re certainly concerned about. Since seat belts are undefeated and saving lives in cars and are by far the smartest thing you can do when you get in your car. And so that recall looks like it’s not going to be going out until right after Independence Day on July 5th.

BMW owners, keep an eye out for that one. And in the meantime, make sure you stay buckled up. Whether you get a, whether your vehicle knows you’re buckled up or not, you should know.

Anthony: Yeah, this day and age, wear a seatbelt, people. Come on. Our other recall this week is from Mazda, potentially 9, 914 vehicles, the 2024 Mazda CX 90.

It is a hybrid electric vehicle. The brakes may activate suddenly due to false detection of certain objects at low speed. Oh, phantom braking? This is not phantom, this is actual braking. Yeah,

Fred: It’s

Michael: phantom in the sense that it’s occurring because it’s not an actual hazard that the vehicle stopping for.

In fact, the vehicle in this case, and it’s at low speed. So we’re not talking about, interstate or even, city road speeds. This is at 10 to 15 miles per hour. The vehicle is essentially detecting it’s its own reflection. Or another vehicle’s reflection and thinking that it’s an approaching object and slamming on the brakes.

Anthony: So it’s a case of vehicle narcissism.

Michael: Except the vehicle doesn’t like how it looks and it slams on the brakes, right?

Anthony: Vehicle body dysmorphism. I don’t know. There you go.

Michael: There you go. And that’s, I think they’ve got a software remedy for that and just a reprogramming, but it doesn’t look like it’s going to be coming out until mid July.

So when you’re driving your Mazda CX 9 near reflective objects at 10 to 15 miles per hour, you want to keep an eye out for this.

Anthony: Doesn’t Mazda also make a model called the Mirage?

Michael: , but they do make the Miata, which is, as we all know, Anthony’s favorite car ever.

Anthony: I’m sure I would love, it’s built for someone of my frame.

I don’t know, Fred, you had something good to say, you’re on mute.

Fred: I was just going to very wittily say that Tesla’s coming out with a new model called Delusion.

Anthony: Ah, that’s a new cologne, Delusion. It’s made by robots. So that’s that’s another episode this week. And what have we learned?

We’ve learned that artificial intelligence is more hype. It’s more hat than cattle. Ha! That Jeff Farah, for the right price, will say whatever. It’s fair to say that. Sure, why not? And that we learned that ABs, while may not be drunk, drive like drunks. Anything else we’ve learned, gentlemen?

That’s it.

Fred: I have the only thing I can add is drive safely, folks. We want to hear you next week. Yes. Drive,

Anthony: put on your seatbelt. Okay. Bye bye. Bye bye. Bye bye. For more information, visit

Fred: www. autosafety. org.