Subscribe to this podcast or subscribe to your car?

Introduction to the Podcast

Anthony: You are 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 Cimino. For over 50 years, the Center for Auto Safety has worked to make cars safer.

Hello America and anywhere else you’re listening to this in the world.

EV vs Gas: Cost Comparison

Anthony: Today is August 27th, and we’re gonna start off talking about a little article from the New York Times that came out today. It was a ev versus gas calculator. It’s basically one of these things saying, Hey, what’s gonna be more cost effective in ice vehicle versus an electric vehicle?

And it has kinda limited choices. It’s very I think even powder puff is too simplistic. It’d be like something for like third graders maybe.

Michael: It’s easy to go click through, right?

Anthony: Yeah. It’s easy to click through, but [00:01:00] I think it, I’m not sure what sense it makes. So I went through and I chose the, just for apples to apples I chose the Chevy Equinox, ev and then the Chevy Equinox gas vehicle to compare them.

And according to them, without any federal subsidy, the EV will be cheaper in five years for cost of ownership. Whereas if I get the federal subsidy, I’m cheaper right off the bat. But there’s a, right before we started recording, Fred was pointing out that he is like they’re missing some data points.

What’s the true here? The true,

Fred: I’m not sure. I compared it to a study that was done by the utility in British Columbia a few years ago. So I looked at small sedan battery powered versus not battery powered. And the comparison that they came up with was quite different than what British Columbia came up with.

So a difference in my approach to it is that British Columbia report showed everything that was behind the [00:02:00] curtain. And this report during the New York Times citing the utility produced by a US laboratory Patel, which is out in Illinois, I think, which is also freely available. But the difference is between those two analyses were pretty startling.

The one that was done by British Columbia showed about a 30% overall advantage in an environment that had 100% hydroelectric power. So I looked at the same. Basic situation in Washington state, which also has a lot of hydro power. And it said that the savings were about four times as big as what their neighbors just to the north had calculated.

So I think we need to look behind the curtain before we put too much credence into this particular utility. It is fun to click around on it though. No doubt about it. [00:03:00]

Hybrid Vehicles: The Missing Piece

Michael: Yeah, and I think one thing that’s really missing from this are, hybrids and plugin hybrids and the cost of them, because I think that the cost would probably be.

A lot closer in the case of a plugin hybrid where somebody was using the electric power to get to work most days. And, driving in essentially like an ev. So maybe make a, this is a kindergarten version. Maybe make a sixth grade version that incorporates all modern cars.

Anthony: It’d be interesting, but I think, is it fair to say that the cost of the electric vehicle should be cheaper over the 15 year lifespan They give? I think so.

Michael: Okay. I think that’s definitely fair to say.

Fred: It should be, but who the hell really knows? There’s a lot of variables involved here that were not included in the study, in particular battery replacement.

So if you look at the need to replace your battery in 10 years, 15 years, something like that, all of a sudden the calculation is skewed far [00:04:00] away from the benefits of the electric vehicle compared to a gas vehicle. ’cause the batteries are very energy intensive. You also got the disposal of the battery.

A lot of factors that are significant that may not have been included in the study, certainly were not visible through the user interface.

Anthony: It’s worth a fun little poke around and I think to your point, Fred, they point out that they’re only calculating out 15 years. 15 years. Why would a car last longer than 15 years?

Get rid of it. Come on. Help the economy. Spend. Now don’t spend, go to auto safety.org and click on donate. Yeah. It’s like spending auto segue.

Fred: But for happiness. You are a master of the segue. Anthony, I’m

Anthony: related to this is an article in Artist Technica titled Study Shows which vehicles pollute the least in every US county.

There was a University of Michigan built a tool that you can use and you can figure out which vehicle is going to burn less CO2 and in compare to saying, Hey, they. Creating a battery electric vehicle produces more [00:05:00] CO2 in the production of it. But then it’s lifetime of use. It produces much less than an ice vehicle.

It’s another fun little tool to poke around. And it turns out, it’s SUVs suck. It’s the quick, that’s the two second takeaway.

Fred: But Hank not good. Not good. Yeah. Whenever you’re driving around 10 tons of metal to go to the store, your local Piggly wiggly to pick up some yogurt, you’re not doing the world any favors.

Anthony: Oh, look at that. Piggly Wiggly mentioned this early on. You’re really trying to, grease the grease, the judges for this Gaslight later on, aren’t you?

Fred: You gotta play to your strengths, Anthony.

Anthony: Oh, I understand. That’s why I don’t play basketball. Let’s continue.

ADAS and Driver Assistance Systems

Anthony: Another RS Technica piece talking about a DAS and humans intervened every nine minutes in AAA test of driver assists.

And this is always the, this is, so this covers our a DAS, our lane keeping, our automatic emergency braking, our adaptive cruise control. And these [00:06:00] are features that I really enjoy, I really like, and I don’t have to intervene so much as I have to be engaged somehow. ’cause if I let go of my steering wheel for more than 60 seconds, 90 seconds, the car’s put your hands back on the wheel.

Come on, do it. Okay.

Michael: So yours, yours would fall into the hands-on category of the study.

Anthony: Yes. It would not be that Tesla Mobile where it’s man, just put a sticky note over the camera. All good.

Michael: Yeah they didn’t have Teslas in that study, but the, I guess the two major modes of, disengagement were when someone was cutting in front of your vehicle from another lane.

The vehicles responses were, all over the place really. It looked like, the it looked like a lot of them were failing to detect the vehicle that had pulled in the lane in front of them. And so the driver was having to, hit the brakes in some cases, maybe a little more [00:07:00] rapidly than they’d want to prevent a collision because the a DAS hat and picked up on the vehicle cutting in.

And the other one was, it looked like they had a lot of. Lane issues like the lane keeping assistance on a number of occasions. If you look at the videos they were taking, it appeared the vehicle was veering towards another vehicle in an opposite lane, and the driver was having to correct it to prevent it from happening.

You know the numbers here, they’re having to intervene. What every. Between every, depending on the issue, like the cutting in looked like it was happening overall every few miles. So it’s not and you think, how often do people actually cut in or cut in front of me in the way that, that, that require response from the vehicle.

And, it, the, it didn’t really get into, how well these systems were performing based on how often those events are happening, but really more over time, how often did the driver have to intervene? So we didn’t really get a good idea [00:08:00] of how the vehicles a DAS systems are functioning so much as how often do drivers need to be intervening to prevent an accident or prevent, some essentially fixing what the A DAS can’t get.

And that’s a lot, I think on average it was, every few miles. So it’s. Basically the entire study says to me, we’re not even close to a point at which humans can or should trust these systems. Totally. And it by any means, and, you’re gonna have to, at least a few times on your commute, intervene to make sure that you don’t crash into another vehicle.

So you can’t truly rely on a DAS.

Fred: The study was based upon traffic in California. Yeah. And intervening every nine minutes, which means basically every 50 yards or so and intervene. But the other thing that’s interesting about that though is that every company who’s [00:09:00] involved in this business talks about the rapid improvements they’re making in their self-driving system and how it’s so much better than it was XX years ago.

Which is interesting because essentially what they’re all admitting is that they’re student drivers. And it seems to me that all these cars therefore should have a clear warning on them. This is danger. Student driver, stay away or danger, incompetent student, driver, stay away. I don’t understand why they’re not doing that.

Anthony: I don’t get it either. My takeaway from this, based off of what you just said, Michael, is not that the a DAA, the eight s systems are the problem. It’s more of drivers of the problem. Because I, whenever I drive on the highway in, in my local area, there’s always some schmuck cutting me off doing something dangerous and whatnot.

And I don’t, the, my adapter cruise control makes an adjustment, but I [00:10:00] have to make an adjustment because they’re not used to people from Connecticut doing 30 miles per hour over the speed limit. Like it’s normal.

Michael: Yeah, I really would’ve liked as another data point in this study to know how often these events were happening.

How often was someone cutting in front of the vehicle so that you could take a look at, how often the vehicle was responding appropriate to the Cutin event. They tracked it over, miles and minutes versus over actual cutin events and how often they worked. So there’s a big variable there that I don’t have any data on.

So it, it makes it difficult to evaluate a few things in this, if they’re dr if you’re driving in a rural area, you might not be cut off. Very frequently at all. And so you’re, the time, the miles and minutes aren’t gonna tell you as much as how often that event occurred versus how much time you were on the [00:11:00] road between interventions.

Fred: All Stacastic events are like that. Think of the weather in, in autumn. It’s generally really beautiful except when you have a hurricane. You say on average, things are great and you develop all of your yard work schedules and all that sort of stuff based upon a continuation of the good weather.

When the hurricanes come, you’re screwed. It’s really the same phenomenon. That we’re talking about here with the the self-driving vehicles, they’re programmed to do things based upon rational behavior and trends that they’ve observed in the past. But these little storms that come through the jerks on the motorcycle, the, the motorcycles I saw on the beltway in Washington, they went by me at 95 miles an hour, all of them on one wheel doing wheelies.

It doesn’t happen very often, but it does happen. And in fact, if you look at any normal drive, there’s always at least one [00:12:00] jerk that causes you to jam on your brakes. Yesterday I had a good one. The woman came out of her driveway in reverse cut me off when I was driving down the road and then stopped in the middle of the road.

This is a, it was a real phenomenon. But it’s not something that’s, I’m sure in anybody’s list of typical events that they’ve programmed their AV to avoid.

Anthony: Thankful here in New York City, everyone drives safe and competently and it’s a bucolic paradise when it comes to driving. They’re

Michael: so lucky.

Yeah.

Anthony: I know. My wife was taking our son to college the other day and some jackass park, his SUV threw a bunch of trash out of his car onto the street and him and his girlfriend started filling up tequila in solo cups in their car while driving. And my wife pointed out them throwing trash out on the street.

He threatened her. My son came out and said, got in his face and picked up the trash and you people are disgusting. And she got a photo of him in his license plate. [00:13:00] And it’s a taxi license plate too. It’s like an Uber or Lyft. So I’m looking forward to requesting him and giving him one star.

Michael: Yeah that’s why we all need to be allowed to conceal carry cruise missiles.

Fred: Ah, cruise missile. They’re not nearly fast enough. They’re Michael. They’re not, they’re gonna go with the side wanderers. Oh,

Anthony: okay. Anyway Fred, do you still have your Subaru? You like your Subaru? I do like my Subaru.

Of course you do ’cause you’re an old hippie.

Subaru’s Driver Attention Systems

Anthony: IIHS has a little study titled Subaru owners see value in Driver Attention Systems quoting from it nearly nine out of 10 drivers who have the feature, which uses a camera to detect signs of distraction and drowsiness told reach researchers that they use it most or all the time, and a majority said they would want the next vehicle they purchased to have it.

They didn’t say how many. These people are being honest though. Quoting further, such a high level of acceptance for a system designed to keep driver’s attention on the road is a bit surprising and very encouraging. Said [00:14:00] IIHS President David Harkey Distraction drowsiness are factors of thousands of crash dusts every year.

So do you have driver focus in your car?

Fred: I do not, no, mine’s a 2020 and I think that’s more recent.

Anthony: Okay. And so this is, these are internal facing cameras that we see in a lot of cars, like the gm super Crew and the Ford Blue crew. They have these, and if the, they’ll use eye tracking and whatnot and see if, hey, if the driver’s not paying attention, disable these features.

And Tesla has one that you can just disable with a sticky note. But this one seems to detect drowsiness and whatnot. Yeah. But do we know what happens when it’s does it do more than just flash up that little coffee cup icon saying, Hey, you need a break?

Fred: It wasn’t described in article, but I think it’s a great thing.

But we were talking last week about the how cars should allow people to defend themselves against the errant behavior of the drivers. One of the things we talked about was [00:15:00] automatic detection of incapacity. And, turning flashing lights on to let people in the neighborhood know that there’s a compromised driver.

This is clear evidence that technology is available now and arguments to the contrary really have no basis in fact.

Michael: Yeah, it’s. It’s pretty clear that the driver focus system Subaru has is analyzing, not just, not just using a camera to look at the driver, but analyzing the vehicle performance data.

It’s analyzing steering patterns and lane deviations to determine when the driver’s awareness is slipped. You could use those same factors to look at the driver’s performance and figure out, are they a drunk driver? Are they having a medical event? Are, a lot of other things.

And that’s one reason why we’ve been advocating for good driver monitoring that can pick up on those things and frankly go further than this system which just warns [00:16:00] someone. If the car is detecting significant deviations from lanes or, terrible steering or any other factors that would lead the car to believe that the driver is a risk to others on the road, the vehicle should be sh shut down, pulled off the si pulled off to the side of the road and left there until, something can happen whether they’re, whether the police are summoned.

Something needs to be, and something needs to be done to prevent drivers who shouldn’t be on the road from driving. And we want cars to be part of that enforcement.

Anthony: This is a feature that it’s not. There’s no regulations coming around the pike for this. There’s no design around it’s not like automatic emergency braking where this has been like, Hey, we’re gonna put some sort of regulations.

This is completely voluntary system that Subaru put in on their cars, on their own volition, right? Yes, apparently. Yeah. So I would like to point this out to every person who’s ever claimed that safety doesn’t sell Subaru disagrees. [00:17:00] ‘Cause this is strictly a safety feature that is no one is asking them to do, and their customers apparently love it.

Michael: Hey, and I would, their, the part of that would be, I would guess that if they did a similar study on a, a different, I don’t know, Subarus drivers I, maybe I just have a soft spot in my heart. Maybe it’s just because of those Subarus love commercials, but they seem to be a more responsible.

Subset of the American population, maybe have you met Fred Perkins? There’s two years. Two years. I’m including Fred in that as a

Fred: I don’t like to be held up as a good example. No, that’s much. Oh, you were actually the bad example of the severe driver. Dressed like working hedge fund.

I appreciate that. Did you said there’s two subset there were, there are two tiers of Subaru drivers. So there’s the sporty tier that likes to get the little zippy cars and zip in and out of traffic and, jump over sand [00:18:00] dunes and all that sort of stuff. And then there’s the weenie population where, I’m a proud member of the weenie population.

There’s a lot of cliches about who actually likes to drive a Subaru, and we won’t get into that here, but I agree with you that generally they’re a pretty conservative bunch of

Michael: drivers. Yeah, no. I look at Subarus and I’ve thought about getting one because, where I choose to live is typically near mountains and areas where during the winter I need all wheel drive or four wheel drive of some type to get in and outta my driveway or get down the road to civilization.

So as right now I’m driving a Volkswagen Jetta that has zero ability to do that. And the closer I get to winter this year, the more I’m thinking about, is this year where I have to pull the trigger and get all wheel drive. And Subaru typically offers that across all their models, which is, definitely attractive.

Fred: Yeah, it works. If you, up here in rural New England, it’s pretty much the state [00:19:00] car of New England. Lots and lots of ’em up here.

Anthony: And Michael, will you keep driver assist on.

Michael: I, I don’t have much in my vehicle right now other than forward collision warning, which is

Anthony: you’re gonna hit that wall

Michael: instead of hitting, instead of actually active activating the brakes.

It just warns me blind spot warning I’ve got as well. But when I get a new vehicle, yes I’m, I talk, we, we have to talk about it so much here that I sometimes feel like I’m at a bit of a deficit because my vehicle, I’m not using it. So I’m not really conversant with the features as much as I should be.

So there’s, I don’t think I’m gonna go all the way and get a Tesla and start blowing around the roads with full self-driving, but, the basic features yeah I think I need to dip my toe in a little more on those.

Anthony: Listeners would you like to help contribute to Michael’s new car so we can talk firsthand about these safety features?

If so, go to auto safety.org, click on donate or add a little [00:20:00] comment. Say this $50,000 is for Michael. It’s personal use only. Legally. I don’t think that will hold up. No. Enter, no,

Michael: don’t do that. But, I’m, I, my car’s about six years old, so I’ve still, it’s running great though. I don’t know.

We will have to see. Okay. My, my number one goal in, in cars is paying as little as possible to get from point A to point B.

Anthony: It’s unfortunate that you have a Volkswagen. Yeah.

Michael: I’ve, I’ve had good luck despite their horrible quality ratings. I’ve had pretty good luck with those guys.

And there’s something about that, that German engineering I can’t get away from.

Anthony: Okay.

Subscription Models in Cars

Anthony: ‘Cause the BBC has an article that VW introduces monthly subscription to increased car power. That’s right. Those who buy an eligible car in its ID three range can choose to pay extra if they wanna unlock the full power of the engine inside the vehicle.

VW says the optional power upgrade will cost 1650 pounds per month or 165 pounds annually, or people can choose to pay [00:21:00] 649 pounds for a lifetime subscription. Okay. So you buy a car and we’re not gonna let you use the whole car. What kind of world are we living in? It’s I bought a Chevy V eight, but I can only afford to use four cylinders.

Yeah, this is insanity. This is

Michael: insane. It’s insanity for a couple reasons. We’ve talked about a very similar thing before. I think very early in their, in our podcasting lives about, I think it was BMW that was putting seat heaters into vehicles and then, and trying to charge subscriptions for them which is Ash Cole

Anthony: give us $3.

Michael: Yeah. And that’s, it’s just a complete departure from the traditional vehicle model of ownership where you buy a vehicle and you own everything that’s already in it. All those options you overpaid for at the dealership are going to work the entire time you have a car. And you’re not gonna be haggled and nagged for subscription money on top of that to keep paying the manufacturer extra over the entire time you own the vehicle.

[00:22:00] But my, I think my real hangup with it is that especially when you start doing things like power upgrades and things, additional vehicle power. I think by definition requires there to be additional vehicle weight in the vehicle. And so you’re putting this additional weight in every VWID three that goes out there, even for owners who are never going to pay you their subscription price.

And essentially it is what this subscription model is doing is adding and if you have, beyond power, if you have, 20 other components in the vehicle that require a subscription but might not be used by other people, you’re just adding needless weight to vehicles, which in crash situations adds needless weight into the crash environment, increasing injuries and making the possibility of a fatality even higher.

So I do not like subscription items on vehicles that add weight to it, I guess is what that comes down to. It seems [00:23:00] wasteful and it seems like it’s just, I don’t know. There’s something that’s really cringey about it to me.

Fred: Michael, I think you’re being small minded about this because, think of all the, think of all the money the automotive manufacturer’s gonna say.

For example, if they put out what could be an SUV, but they require a subscription fee to use the backseat, then they can call it a sedan, right? Because it’s, or a sporty sedan. And if you happen to have a lot of children unexpectedly, you can just go ahead and pay that fee to use the backseat. I think it’s a great idea.

I think I can see all kinds of possibilities for companies to save investment and different models. And for pickup truck, if you want to use the rear bed, you have to pay a $5 fee. It’s a lot cheaper than going out and renting your own pickup truck for that yard cleanup. You gotta do.

Anthony: I don’t plan on ever using the airbags in my car.

Okay. So I shouldn’t have to pay for them unless they’re deployed. [00:24:00] That’s right. Okay.

Michael: Yeah, and that, that raises an issue. Obviously I don’t think anyone is offering subscriptions for safety systems yet. Although, arguably in the past, manufacturers have, we’ve seen things like automatic emergency braking before it’s required.

They can be sold as options, they’re not standard yet. So no safety features should be, should come at the cost of a subscription, obviously. And I think there were, there’s plenty of law even in this crazy iteration of the United States to prevent manufacturers from requiring folks to pay money for safety systems that are required by federal regulation.

But for all of those things that could help with safety, like right now, some of those distraction features the Subaru feature that we were just talking about, they could be offering that at a, as a subscription feature. And they’re choosing not to, which is good on them. But that’s where our concern really lies, [00:25:00] is preventing safety.

Equipment from becoming a subscription feature that’s really only available to a certain class of people that can afford those subscriptions. Are you familiar with healthcare in this country? It does sound a lot like that, doesn’t it?

Anthony: So I like this. So they’re adding unnecessary weight to a vehicle.

And if I wanna actually use which will which will cause the car of tires to wear out longer, I will not be able to access these features unless I give them more money. And it just goes back to the question I think I asked you guys regularly is what did, what did I buy?

Michael: Did you buy anything?

Did

Anthony: I buy this? Do I own this vehicle? Because apparently, according to Volkswagen, I only own some of it, but I’m sure Volkswagen will figure this out. Volkswagen remember Dieselgate ha farik newgen. I know what it means now.

Michael: Yeah. Volkswagen may not be around by the time I’m ready to buy another car, so

Anthony: they might not be around in the next, by the time this airs.

Oh God. Okay, let’s stick with Europe [00:26:00] here.

European vs US Vehicle Safety Standards

Anthony: An article from MSN titled, unfair Competition Worries about European Road Safety After eu, US Trade Agreement, quoting from the article we now have in the EU technology such as automatic emergency braking, pedestrian Protection Systems, and Lane Keeping Assist Systems explains Dudley Curtis, ETSC, the European TRA Transport Safety Council Communications Director.

These are just three examples of a technology that are mandatory in the US but not, I’m sorry, mandatory in Europe, but not mandatory in the us. And so I guess the issue is that European cars, by default have to become safer than what they sell in the us. Oh yeah. So are the European car manufacturers saying what we’re gonna export dumber cars, less safe cars to the us?

Is that

Michael: No. It is the opposite. I think here they’re gonna have the EU is going to have to accept. Vehicles from the United States that don’t meet the European standards. And we’ve spoken ad nauseum about the [00:27:00] how Europe has raced ahead and vehicle safety. And the numbers prove it.

But in this case. The United States actually might stand to gain from a safety perspective because the vehicles that could, would be shipped to us could have the EU technologies on them. ’cause it’s kind. I think this one, there was another agreement that was reached with Japan, I believe, as part of the tariff issue that basically allowed the United States to export vehicles to Japan and not have to meet Japan safety standards, which essentially downgrades the fleet on Japan’s roads.

This, in this case, I think this is a bilateral agreement where we are shipping vehicles to Europe, but also Europe is shipping vehicles to us. And in America we stand to gain because our safety standards are lower. We don’t have intelligence, speed assistance. They’ve had, although automatic emergency braking, which is.

S this is all still hung up [00:28:00] in court for the moment. It’s supposed to be available in all American vehicles by 2029. Europeans have had it for years now, and those vehicles are gonna be in imported in America. And our, our, the safety quality of the vehicles going on the road in America that are coming from Europe will be higher.

Fred: I’m just thinking that I, it seems like making America great again is a zero sum game, which can be won by making Europe and Japan crappier. So I guess it’s a comparative event, right?

Anthony: For

Fred: Fred’s,

Anthony: Positivity podcast. Oh, God. Okay.

Tesla and Robo-Taxi Market Predictions

Anthony: We have a bunch of Tesla nonsense to get to. And I think before we even do that, it’s time to let’s take a deep breath stretch and go into some gaslights.

I’m gonna kick us off ’cause my, I have multiple gaslight choices. It’s just but my first one’s related to Tesla. So [00:29:00] here it goes. Arc investments. That’s right. Going after arc. Quoting from ARC Investments. According to our research, Tesla’s Robo Taxii business could represent roughly 90% of its enterprise value.

By 2029, capturing a significant share of arcs projected $10 trillion Global robo taxii market. So these clowns think by 2029, the robo taxii market will be worth $10 trillion globally. Why does anybody give these people money? Like it’s go what? Either that, maybe there’s something that they’re not sharing with everybody where the inflation is gonna be so rampant that $10 trillion makes sense.

That’s the only way this makes sense. Park investment. My Gaslight of the week or. Can I add on another?

Michael: Yeah, please, dude.

Anthony: Okay. My other one is the CEO of Uber whose name I will totally scramble the shore, the whatever,

CEO of Uber, who said [00:30:00] Alphabet’s. Waymo has already shown it’s possible for self-driving vehicles to surpass human abilities.

Did I steal yours, Michael?

Michael: Yeah.

Anthony: Ah okay. You gave the point. Point through. You

Michael: can go through the CEO of Uber part. He’s not my guy.

Anthony: Oh. See, that’s that guy claiming that Waymo’s already shown it’s possible for self-driving cars to suppress human abilities, and that is not true.

Michael: Yeah. Yeah, he’s, he is over touting way Waymo there, but he Musk responded to him, and that was my light, that was my gaslight of the way.

’cause take

Anthony: it away.

Michael: Musk says, LIDAR and radar reduce safety due to sensor contention now, and in a different reality from what Mus lives in, we would, we might call that sensor fusion, where you have multiple sensors that are capable of detecting things. And the thing that we and he says, I’ll go on with what Musk says.

If lidars or radars disagree with cameras, which one wins? This? Sensor ambiguity causes increased, not decreased risk. [00:31:00] That’s why Waymo’s can’t drive on highways. He said that last sentence is total bullshit. The reason Waymo’s can’t. Or aren’t driving on highways is because they’ve chosen a different path.

They’re actually focused on getting things right in the city and in the areas where driving is particularly dangerous and particularly difficult for autonomy, Musk’s company actually chose to start out with vehicles that were supposed to stay on divided lane highways, where we have the lease fatalities and injuries.

And where cameras can, systems can probably operate with, a sufficient margin of safety in certain circumstances. Not in low visibility though. But the problem is, it seems like this guy just doesn’t have an understanding of some of the basic concepts here.

Anthony: Wait, Elon doesn’t understand something.

You’re outta your fucking mind, you’re my gaslight of the week, Elon.

Michael: It’s just not that simple a proposition.

Sensor Integration Challenges in Autonomous Vehicles

Michael: Yes, there’s going to be different answers that are given to you [00:32:00] if you’ve got lidar, radar and cameras and other sensors on the vehicle. But the whole job of the engineers behind that and software engineers and all sorts of folks who are working is to build a system that brings the best of those into play in the areas where they are able to work the best.

Obviously the cameras that are loaded onto Teslas are not going to work well in low light conditions or in, in, in conditions where, visibility is restricted and there’s no way around that. Using cameras, neural network, whatever Musk says in that area is just stupid. It’s, there’s no other way to put it.

You, you’ve got Michael somehow integrate other sensors in and figure out a way to develop, develop software that can distinguish the, distinguish between the different signals you’re getting from a Lidar and a radar [00:33:00] and all these other types of sensors to figure out the path of the vehicle depending on what the circumstances are.

There’s a lot to that and it’s not nearly as simple as Musk wants us to have us to believe. So we’ll buy his all camera system.

Fred: What Musk is basically saying is that common filters don’t exist and can’t work. Common filters integrate various sensor inputs and. Calculate the preferred trajectory of vehicles.

They work for nuclear submarines navigating onto the ice in the Arctic, for example. They work on airplanes. They what Elon is saying is the technology that we use to drive nuclear submarines and military vehicles cannot be applicable to the task of driving a vehicle which is pure bullshit.

Human Sensory Integration vs. Autonomous Systems

Fred: And I also wanna point out that in reality human beings integrate lots of inputs. In fact, if you think of it for a moment, we even integrate in [00:34:00] inputs from our butt when we’re driving. Because when you’re driving down the road, that’s gravelly and crumbly, right? Like a dirt road. The sensations you get from your butt are quite different than the sensations you get from driving on a well-maintained highway.

So your mind reacts to your, but. And picks a different mode of control based upon, all those various inputs.

Critique of Elon Musk’s Approach to Self-Driving Technology

Fred: So yeah, Elon is just blowing smoke up people’s asses once again. Seems to work for him. He is gotten a lot of money outta that, but people apparently love it. Yeah.

Anthony: Yeah. The mind connection is highly underrated as a driving style.

I think so. I think so.

Fred: Even our president, no, nevermind. Yeah. He

Anthony: no but Musk, you’re right, saying what do we do in these systems, conflict and whatnot. As you pointed out, Fred, like there, this is software 1 0 1, where on very simple systems, you go, oh, we have two input values, and they disagree.

You put in rules that say this is the [00:35:00] one that, this is the primary source we’re gonna believe. This one with air airline commercial airliners, you could be in a jet plane and you could be reading a wind in speed indicator of 500 miles per hour, but the GPS says, Hey, you’re not moving. Why is that? I dunno.

’cause you’re running into a 500 mile per hour headwind. You’re still flying, but you’re not going anywhere. So which system do you do this? They have very basic rules. It’s I oversimplified it, but this is not a magic problem. This is not something difficult. Instead, this approach is saying we only have one system.

We just have to trust the cameras.

Michael: That’s just total bullshit. We do it all the humans, time as humans. What you doing when the cameras suck Humans, right? We are driving down the road and then the road looks clear ahead of us and nothing’s going on, but then we hear a siren. And immediately we have to change what we’re doing.

And even though the road’s clear ahead of us, when we recognize that there is a siren behind us, we know that we need to, stop [00:36:00] moving and get over to the right lane or onto the shoulder to get outta the way of an emergency vehicle. And that’s two different sensors coming together. And the human mind, de deciding which one needs to take precedence over the other to make a safe decision.

It’s not, it’s not rocket science. And it’s something that makes, the ability to combine those two sensors in the human mind is analogous to using different sensors in a vehicle and can produce better safety outcomes apparently, that doesn’t exist in Musk’s world.

Fred: All right. So Fred, my Mike Gaslight turn. Okay. I’m going with Chad, GPT ha.

Chat GPT’s Take on Self-Driving Car Safety

Fred: For for a couple of reasons. I’m also gonna do the tau on artificial intelligence, but this is the gaslight part of it. So I asked chat, GPT are self-driving vehicles safe? Are self-driving cars safe? And what chat GPT said is, that’s a great [00:37:00] question and the answer is a bit nuanced.

One’s a great question and you’re so good

Anthony: looking.

Fred: Yeah, but the way it summarized it is it said potential safety benefits. What we know today in 2025, human error, distraction, speeding, drunk driving fatigue is responsible for over 90% of crashes. In theory, autonomous vehicles can reduce these risks since they don’t get tired, drunk or text while driving.

AVS can react faster than humans in some situations. EEG automatic emergency braking. So that summary is based upon an urban myth that has no basis in research or fact. It became a popular urban myth when the previous Secretary of Transportation decided to misinterpret a study that Nitsa had done several years earlier.

I, as our listeners, no doubt know by now, there is no [00:38:00] data to support that idea that human errors responsible for over 90%. In fact, 94% is the number of people generally use of crashes. People are a factor, but in fact, what’s true is that humans are a factor in a hundred percent of automobile crashes because automobiles are built by humans.

So if there were no humans, there’d be no crashes, right? So anyway, chat, GPT pushes us down that road pretty far. And so giving it without context and without saying what safety even means, I’m given the Gaslight Award to chat GPT this week.

Anthony: Pretty good. All right. The judges have decided they’ve done their homework and Michael is the winner this week.

Oh, damn it. We all jumped in to Michael’s gaslight. We all got passionate about it. It was good. So that’s [00:39:00] so no passion

Fred: about chat. GPT Come on Anthony. I, we spend all our passion,

Anthony: On Michael’s Tesla thing coming. You, Joey did too. You know what

Michael: we should be doing is feeding our Gaslight of the week into chat, PT, GPT and let it decide who lives.

I quit. Alright, let’s let chat. GBT take Anthony’s job.

Fred: Alright, let’s do that. All right. I don’t think Che GPT can wear those attractive t-shirts though.

Anthony: Yeah.

Fred: I haven’t figured that out yet.

Anthony: They can’t. You listeners, you’re so lucky. You’re, this is not a visual podcast anyway. Fred, you’re saying your tau this week is gonna be on chat.

GPT, is that right? Artificial intelligence?

Fred: Almost.

Anthony: It’s on it’s coming up next. Is everybody ready for it? I know we are, huh?

Fred: It’s on chat GPT five.

Inconsistencies in AI Responses

Fred: So I did, I asked that same question chat GPT five and it gave me a different answer. So I and the relevance here is that people are touting the use [00:40:00] of AI to drive vehicles.

In fact, Tesla’s gone so far down the road to say they’re using an end to end. AI to drive all their vehicles now on full self-driving. So inquiry minds wonder since AI gives inconsistent results with the same data how does this really work? Anyhow, it’s what chat GPT five said is that self-driving cars in 2025 are safer in some ways, but still face challengers.

And then they go on to talk about safety improvements and controlled environments. Autonomous vehicles have 40% fewer accidents than human driven cars, thanks to advanced sensors. Where’s that data from? That data is from actually they gave us a citation that’s from motor watt. Com motor. What the hell is that?

Not motor. Yeah. I’ll

Anthony: send this link to you. Chat. GBT invented that [00:41:00] website.

Fred: Yeah, but it was written by Alex Rory self-described as an EV market expert. Oh, wow. Oh wow. Yeah. Yeah. That’s the citation they provided there. Go. Great. So remember that number right? And controlled environments have 40% fewer accidents.

But if you continue scrolling down their analysis, what it says later on is that in 2024, the number of self-driving car accidents nearly doubled. And similarly, what semi-autonomous vehicle accidents increased 35% year over year. So that doesn’t sound too safe to me. And then it all goes on to say self-driving cars are more than twice as likely to be involved in an accident compared to human-driven cars.

So within one result, from chat GPT five. They’re saying both. They’re safer and they’re involved in twice the number of accidents. A lot of other interesting stuff in here. So they actually do produce citations and some of [00:42:00] the information they’ve got is actually correct, but it seems to be correct by accident.

Good illustration of the difference between correlation and causality, but clearly it’s based on headlines and information that is just bullshit. So they never tell me what safe means and overall chat GT five, I’m giving it a ninth grade social studies teacher grade of B minus with a red pencil mark that says, good effort.

You wanna encourage your students, right? But internal conflicts exist. And you really need to stick to primary sources.

Michael: Yes.

Fred: Again, the relevance I want to point out is that this technology with internal inconsistencies and no basis in primary sources is what people are now using to drive their cars down the road.

And your life depends upon the accuracy and verity of these AI [00:43:00] tools. So abandoned hope all ye you winter here or Tesla, whichever you’re using. Thank you.

Anthony: While you were doing that, I went into Google’s ai, Gemini typed in our self-driving car safer than humans. And it said it’s a complex one but then it basically says data from various sources including nitsa reveal that self-driving cars.

Particularly those operated by companies like Waymo, which Google owns are involved in fewer injury causing crass and fatal accidents per million miles driven compared to the national average. What that’s a little what, A little funny, but it’s funny ’cause it’s citing things are law firms it’s citations are like ambulance chasing law firms and things of that nature.

And some sources I’ve never heard of before. Something called Bonfire Leadership Solutions. It’s, it’s not good. It’s not, yeah. The [00:44:00] sources that are

Michael: response, the sources that both chat, GPT and Jim and I are relying on today, and both of your examples are terrible.

Anthony: Yeah. This is, personal injury lawyers, if I

Michael: source, if I put those kind of sources, in a footnote, in, in any type of argument, if it was a letter to Nitza or part of a lawsuit, I would be laughed out of the room.

So I guess that’s what I think of chat, GPT and Gemini at this point.

Fred: There’s another citation by chat, GPT five that says, companies like Waymo report 57% fewer police report crashes. Which is true and very misleading because when they did the analysis by Swiss Reed, they threw out all of the crashes that were reported in the nitsa standing general order.

Yeah. Anyway, that was interesting. And that was done by a company called Consumer Shield, written by Sarah Edwards, and actually reviewed by Edward Ramirez, [00:45:00] who is one of the owners and principles in Consumer Shield. And they go on to say, consumer Shields content contributors, which is like Sarah, have demonstrated legal expertise.

Most have graduated from law school, some have passed the bar exam, and any who don’t have a JD instead have credentials in another relevant area. For example, a master of science and library and information science or certification in a field like tax or real estate. So this is the expert source being cited by chat GPT five and their analysis of automotive vehicle autonomous vehicle safety.

So yeah it’s just bullshit being elevated to the level of truth. It’s. Like certain politicians who have now achieved the apotheosis of vulgarity and have national offices based upon that.

Anthony: Brad, I’m gonna tell you something right now that’s really gonna bother you and upset you. [00:46:00] Oh, if you used this as your gaslight, you would’ve won.

See the

Fred: gaslight is too narrow now I just not gaslighting

Anthony: you. Oh, come on.

Fred: Alright, listeners, let’s we go back to reality. You l you Anthony. Back

Anthony: to, let’s go back to the world of Tesla. And this is a good one.

Legal and Safety Issues with Tesla’s Self-Driving Claims

Anthony: It’s from electric titled Tesla Loses Bid to Kill Class Action Over Misleading Customers on self-driving capabilities for years.

Everyone crack open the champagne. Yeah. Quoting a judge has ruled to allow a class action lawsuit against Tesla over claims that the company has been misleading customers about self-driving capabilities for years. We’ve been talking about that for years Further. Tesla claimed that all vehicles built since 2016 have the hardware capable of achieving full self-driving, which isn’t the case.

And Tesla has been selling a software package called full self-driving that it claimed would deliver unsupervised level four to five self-driving. And it [00:47:00] hasn’t. We’ve been saying this for years, ladies and gentlemen, it’s about time the legal system listen to us. Eventually the Supreme Court will say they can do whatever the fuck they want, but for right now, it feels good.

Michael, as our resident legal historian,

Michael: This is, this lawsuit is filed, I think quite some time ago. Tesla’s been, doing their typical delay, deny. It’s an interesting case because. Tesla doesn’t use traditional advertising. You don’t see Tesla commercials on TV during the Super Bowl and you don’t see billboards with Teslas on them, and they have a very different philosophy.

And so their marketing and advertising typically is, goes through their fanboys on the internet. But through Twitter, through blog posts, it’s website, things that Elon says, that type of thing. Elon gets so much of a attention that what comes out of his mouth is as good as an advertisement for [00:48:00] a lot of folks.

So it’s an interesting, it’s gonna be an interesting case mainly because of, the traditionally when you see. A case based on, false advertising or claims that are similar to fraud. It would, be supported by a lot of advertising television advertising, print advertising, that, that was making claims that don’t align with reality.

So that’s the interesting part of this, but their website made

Anthony: these claims too, didn’t it? Yeah. Yeah. So that’s advertising.

Michael: So it’s an interesting case and there’s a couple of cl there’s a couple of issues with it as well. The, a lot of people who bought Teslas, if you buy a Tesla, I think the first thing you should do is send the letter to Tesla to opt out of their arbitration agreement.

They don’t let you do that electronically. They want to make you send it in. It’s like the rebate process where you have to have your receipt, you send in your receipt to the manufacturer to, and you have to make sure everything’s perfect or they won’t give [00:49:00] you your money. In this case, you’re actually, by opting out, you’re actually allowing yourself to avail.

You’re availing yourself the ability to go to court later if something happens and not to be stuck in arbitration or the kangaroo court. And you’ll see that’s true in this case because the class action that was certified here, covers, it covers either California residents who purchased full self-driving in a very small window between October, 2016 and May, 2017.

And then the second class covers full self-driving owners who opted out of Tesla’s arbitration agreement between 2017 and 2024. So if you didn’t send your letter in or opt out of Tesla’s arbitration agreement in the form they required you to in 2017, which I believe was slightly different, you’re not covered by this class action.

I guess this is, my legal or lawyerly grandfatherly advice here, always check the [00:50:00] contract you’re signing and if there’s, some contracts of adhesion, you’re not going to be able to get out of the arbitration provisions. You have that option with Tesla and you have it with a lot of other manufacturers who you’re making these agreements with.

Opt out of arbitration. Look through your big boring contract that you’re signing that you really don’t have, you can’t modify. You have no choice in the contract, but if there’s any way to opt out of arbitration, do it as soon as possible and follow the directions for doing so. It might make your life better in the long term.

Anthony: Another Tesla related story.

Tesla’s Delayed Accident Reporting

Anthony: Tesla faces new federal probe into delayed accident reporting. And I think we briefly mentioned this last week where it, I’ll quote from this, in a federal filing released last Wednesday, the National Highway Traffic Safety Administration said Tesla may have violated reporting requirements that mandate automakers disclose crash data within days of an incident.

Regulators found numerous incident reports in which Tesla submitted information months [00:51:00] late, often invoked filings rather than on a rolling basis. So their automatic, their ADAS systems. Crash and then they’re just like, keep it cool, bro. Keep it cool. We’ll just slip it in like at 4 59 on a Friday before a holiday weekend.

Is that, and NSA’s no

Michael: it’s, it’s hard to tell. NSA doesn’t provide a lot of detail here. It’s a little a general overview of what they’re looking into, but essentially what it comes down to is Tesla’s collecting the best crash data of any manufacturer out there when these systems are in operation.

So there’s absolutely no excuse for Tesla not to be complying with the reporting provisions of the standing general order. But the problem here, and it’s, the first question Anthony asked me when he saw this article is, what is the punishment? The punishment here is. Paltry in relation to Tesla’s wealth.

It’s a 20, I think the maximum [00:52:00] daily fine under the standing of general order is now a little over $27,000 per day. Frankly, Tesla could afford to not report any of its crashes to the standing General order pretty easily with the market capital. And, all the money that, that Tesla has these types of fines just simply aren’t enough to incentivize accurate reporting.

We think that the fines, not just for standing general order reporting, but also for communications between NSA and Tesla that are in safety investigations. NSA sends out information request. To Tesla, when they’ve got problems with their summon mode or their full self-driving.

And Tesla, from our perspective, provides a la they don’t provide full responses to nits and Nitsa has to go back to them multiple times to get full responses. I wish Nitsa was a little stronger there and just [00:53:00] started finding Tesla. But even if they did, the amounts that they can find manufacturers under the Safety Act simply aren’t high enough to make those manufacturers really respect the authority of the government here.

They can afford to pay the low figures that are allowed. So I guess this is my pitch, just to dramatically increase the civil penalties that are available to Nitsa when they’re trying to get information from manufacturers.

Anthony: So let’s add two zeros to that find. How does that sound?

Yeah. Okay.

The Safest Seat in the Car

Anthony: Before we jump into recalls this was a good article we came across where it helps me feel better about my childhood. See, I was the youngest in my family. I have two older brothers, and we’d get into the family car and being the youngest, I’d have to be like, I got the hump. You have to sit in the middle of the back seat over the hump or the transmission line, transmission and the gears.

And I don’t know what, I don’t know what was in that hump. [00:54:00] Anyway from an article in Pop ps quoting the rear seat, middle position would be the safest because you’re furthest away from side impact intrusion or penetration. So all these years I thought getting crammed on top of the hump was punishment, but they were all trying to keep me alive.

Michael: This is a really complex question and I’d like to give a shout out to Byron Block, who is an old friend of the centers who was the main interviewee in this article, and who did a really good job of trying to explain a really complex sub subject because. Everything comes down to what type of crash you’re in, statistically, that center rear driver’s seat may be the safest overall, that center rear driver’s seat also Passeng doesn’t Passenger seat the center rear.

Yeah, the center rear passenger seat doesn’t have the advantage of being protected by, frontal impact airbags, and as we’ve spoken about, the driver’s sheet is probably, or traditionally been one of the most protected spaces in a car because [00:55:00] statistically there’s al always someone in the driver’s seat.

And there’s certainly the rearer seat’s been neglected over time because it’s rare for people to be in the rear seat than it is in the driver of the passenger seat up front. And it’s a very complex subject because. You just can’t, children need to go in the rear and not the front.

And you’ve got to consider a lot of factors. One of them in particular, an article that Byron says, you want all the side windows and the sunroof to be laminated glass because tempered glass immediately shatters into thousands of little pebbles of glass. That’s a hundred percent true.

From the perspective of crashworthiness, you want alt laminated glass in those areas. Laminated glass will remain structurally intact to some extent, and can even help preserve the cabin space that you have and adds to the strength of the vehicle. Helps the roof be a little stronger if it stays together.

However. After the [00:56:00] crash, when you have to worry about escaping the vehicle, that laminated glass then turns into a barrier to exit if you’re trying to, if you’re submerged or if there’s a fire and you have to get outta the vehicle quick. So there’s a trade off there as well. And it’s, I guess all of this is to say that it’s a very complex subject about which of the safest seat in the car.

I think you might, the it’s a question that, that isn’t proper. You need to ask the question in a side impact collision from this side of the vehicle. What’s the safest seat in the car in a rollover collision? What’s the safest seat in the car? Because it’s really hard to say just overall what the safest seat in the vehicle is without considering the type of collision that you’re looking at.

Anthony: It’s a fun fun little read and with that, something that’s not complex.

Recent Vehicle Recalls

Anthony: His recalls. Let’s start off with, oh boy. Chrysler 78,989 vehicles. The 20 [00:57:00] 24 20 25 Jeep Wrangler. Oh. Warning. This is tire pressure. The tire pressure center sensor is unable to warn a driver if the pressure is outside of the recommended limits and can cause a vehicle crash without prior warning.

Wait. So if the tire pressure monitoring sensor is not working, how would that lead to a crash? Like your tires are overinflated. That’s more in danger. Is it under, what

Michael: if your tire was experiencing a deflation and the tire pressure monitor couldn’t pick it up? It was just Oh, I

Anthony: get it. Okay. Alright. Get that

Michael: fixed.

Yeah, and it’s an, it’s weird, isn’t you? There’s, it’s an odd reason for this, it’s almost like. This is a remote start antenna cable. So that’s, I’m assuming the cable that when you hit the remote start button on your vehicle, the antenna picks up that signal and starts your vehicle.

How that is somehow connected to the tire pressure monitoring [00:58:00] system, I don’t know. But it seems like another case of two different systems interfering with each other. But there, it doesn’t appear to be a software issue because they’re fixed for it is to replace the remote start antenna cable. So still a little mystery to this one, but owners should be hearing from Jeep in early September.

So in a couple weeks here, or one week

Anthony: next

Michael: week guys.

Anthony: So exciting. Get it done for Labor Day. Next up Mercedes. 3,749 vehicles. The 2024 to 2025 Mercedes-Benz, A-M-G-G-L C3 four. Oh God. A bunch of, you got a Mercedes? Yeah. Don’t even, it’s, it ends with A GLC in it. It’s got something, there’s a whole list of these things.

Michael: Yeah. If you’re a Mercedes owner listening to the podcast, you’re gonna have to get a NSA’s website and check out the recalls. ’cause we’re just, we’re not reading all that nonsense. No

Anthony: better yet. If you’re a Mercedes owner and you’re listening to this podcast, go to auto safety.org and collect, [00:59:00] donate.

You got the coin. Come on. It continues down to EQE vehicles. And what is going on here? A bolt that secures the steering coupling may not be torqued to production specifications. Oh no. And you’ll the steer steering rack may loosen over time. Such an occurrence can result in loss of steering control.

See, I understand how that increases risk of crash. That’s a, what were you saying about German manufacturing earlier, Michael?

Michael: I would maintain that, especially if we’re talking about Mercedes. They seem to have a, they seem to have things pretty dialed in and on their safety systems and they, yeah.

From, from reading through thousands of these over the year, Mercedes seems to be one of the most responsive corporations to safety issues that their owners are having so hard to criticize them. But in this case, they really needed to get that torque a little better on these steering coupling.

And the steering rack owners are gonna hear about this in about a month and a half, mid-October, and they’re gonna have to go in and [01:00:00] get some new screws and nuts put into their steering system.

Anthony: Alright, last recall General Motors 23,656 vehicles. The 2023 Corvette, the 2024 Corvette, the 2025 Corvette, the 2026 Corvette.

Michael, you gotta get your Corvette. Then

Michael: Corvettes all four of them. Yeah. I’ve got one from each of those model years, all pink. But I’m not covered by this recall ’cause it only covers vehicles with a left side radiator fan combination. I’m, I’m a right side guy. Wait, what’s, what is going on here?

There’s essentially when you are filling up your filling, upt your car with gasoline and you have a little extra fuel and it spills into the fuel filler for whatever reason the models of these Corvettes that have a left side radiator fan combination. Risk catching on fire. My,

Anthony: my [01:01:00] toupee manufacturer said I’m not allowed to fill my own cars.

’cause the fuels could the vapors could set my tope on fire.

Michael: Yeah, that’s because your toupee is cheap and not made with human hair. It’s made with polyester, dammit. The, and this is, I don’t know if there’s, they say they’ve had four incidents of this happening. A lot of these came to GM through social media.

So the power of social media for safety is on display here. And GM took a look at it and said, you know what? We’ve got a problem here. And I would advise, for all of you age to older men who are, experiencing the desire to have a Corvette, I would say be really careful when you’re pumping until mid-October.

Anthony: That’s the same thing that’s you’re gonna, a cardiologist notified.

Michael: Yeah. That’s when you’re gonna be notified of this recall. And I don’t know, they, it looks like they’re gonna install a shield to divert any spilled fuel. But [01:02:00] between now and then, don’t spill your fuel. Really focus on getting the pump into the tank and leaving it there until operation is complete.

This is,

Anthony: oh my God,

Fred: Michael, I think that was a very cold comment about Anthony’s toupee. And I want to point out that there’s different grades of polyester, and his toupee is definitely a higher grade polyester.

Anthony: Good for you. Oh my God, my midlife crisis doesn’t involve a Corvette. Maybe it should. I don’t know.

Conclusion and Call to Action

Anthony: Hey, with that, folks, that’s another episode down the drain. Go to auto safety.org, click donate. Tell all your friends send us pictures of you, your Corvette, and your girlfriend that is way too young for you. And we’ll just have a giggle little next time. Bye-bye. Thank you. Bye-bye.

Fred: For more information, visit www.auto safety.org.