My name is Avi Kelin. I'm an attorney with the law firm of Genova Burns in Newark, New Jersey. I'm the Chair of our firm's Autonomous Vehicle Law Practice Group. And today, we're gonna focus on a CLE entitled, Autonomous Vehicles Legal Considerations. The goal of today's seminar is to focus on a few topics. We're gonna consider what are autonomous vehicles, also known as self-driving cars, there is a few different names for it. What is the current state of the technology and where is the technology going? We're gonna consider the benefits or potential benefits of widespread adoption of this technology. And then we're gonna shift gears a little bit and focus on the current state of the law. Some speculation informed speculation on where the law may be going. We're also gonna consider open legal issues. And then we're gonna spend a few minutes considering what type of industry, and what type of government response maybe needed to prepare for widespread autonomous vehicles. I'd actually like to start going back to the year 1957, and there was a great advertisement in magazines and newspapers across the country, sponsored by the electricity industry at the time. And there's an image of a family of four sitting in a convertible, the parents in the front, the kids in the back, and they're all playing a game in the back. No one's facing the front seat. No one's actually controlling the car and the message of this advertisement said electricity may be the driver. One day your car may speed along an electric super highway it's speed and steering automatically controlled by electronic devices embedded in the road. Highways will be made safe by electricity, no traffic jam, no collisions, no driver fatigue. And again, this advertisement ran in 1957. We've come a long way since then. But tech technological advancements have made this potential vision from 1957, where it brings us on the brink of reality. So what are autonomous vehicles? It's actually not a simple question, or I should say not a simple answer. The National Highway Traffic Safety Administration recognizes six levels of autonomy. Level Zero is the human driver does everything. And you can think about a Model T, cars 50 years ago, a hundred years ago where basically there was no safety features. The car was just a piece of equipment, and the human driver does everything. Level One is getting a bit more advanced. That's defined as an automated system on the vehicle that can sometimes assist the human driver conduct some part of the driving task. That could be things like cruise control, lane guidance, things that we have in our vehicles now, where the vehicle itself actually assists the human driver who's still in control, by the way, but the vehicle assists the human driver conducts some parts of the driving task. You can also think of this as backup cameras or other cameras that you have on the car, where the car is pulling a bit more weight than on Level Zero, where the human driver does everything. Moving up the chain, Level Two is defined as there being an automated system on the vehicle that can actually conduct some parts of the driving task while the human continues to monitor the driving environment and performs the rest of the driving task. So in this Level Two, we have a vehicle that actually can itself do some parts of the driving. However, it can't do all driving and the human needs to be available to monitor and be able to take over when necessary. Continuing to move up the chain, Level Three is defined as an automated system that can both actually conduct some parts of the driving task and monitor the driving environment in some instances, but the human driver must be ready to take back control when the automated system requests. So the difference here is that the vehicle really can do it all in some limited circumstances, in some environments, in some instances. That depends on the driving environment, though. It's not gonna be able to work in all environments. So the human driver, again, must be ready to take back control when notified by the automated system. Level Four that's defined as an automated system that can conduct the driving task and monitor the driving environment. And the human does not need to take back control. So this is really, the car does everything. The vehicle does everything, but the limitation here is that the automated system can operate only in certain environments and under certain conditions. So this is not the case where if I'm driving cross country and there's inclement weather, and there's other factors involved, it doesn't necessarily mean that I'm gonna be able to sit back and watch a movie or work on my laptop the entire way, but in some circumstances, in some conditions, the automated system's really gonna do all aspects of this driving. And in those defined environments, I, as a human driver, I'm not gonna have to take over control and not gonna have to operate the vehicle, but again, that's not gonna be the case in all environments. So, we can draw a distinction, maybe the vehicle is set up for highway driving, but not local roads, in that case, the vehicle can really drive itself on the highway, but I need to do the first mile to get onto the highway. And the last mile, once I arrive at my city destination, but once it's on the highway, the vehicle's good to go, and I don't need to even be sitting at the steering wheel at that point because I'm gonna be notified when we're done with highway driving and I'll take over at that time. Level Five, which is the highest level is defined as an automated system that can perform all driving tasks under all conditions that a human driver can perform them. So that's really, you can envision in that scenario, this Level Five, there's no steering wheel. There doesn't have to be mirrors or even windshields because the human driver is not gonna be driving at all. It's gonna be the vehicle that's performing all driving tasks in all circumstances. So that's the theoretical levels that have been defined as an industry standard. But let's take a moment and consider what the current state of the technology is and where it may be going. So as far back as 2013, there was an auto critic for the New York Times who reported driving an INFINITI Sedan on surprisingly long highway stretches without touching the accelerator, brake pedal, or steering wheel. So, again, this is going back nine years from when we're recording, but back in 2013, it was surprising to people that even on a highway, the car would basically control itself. There are ways for it to keep a safe distance from cars in front of it. There are ways for the car to stay in its lane. And basically, I didn't have to do anything if I was driving that car back in 2013. I had to get on the highway, but I can go a few miles at a time, and the car would basically control itself, until I'm ready to exit. What's going on now is what we have, perhaps the best-known example of the Level Two car is Tesla's Autopilot feature. So Tesla cars can accelerate, maintain lane position, change lanes and park without any driver input. However, the reason why this is Level Two and not higher is that Tesla requires drivers to keep their hands on the wheel and eyes on the road at all times. So that a human is ready to take over when the system notifies you. But really, there are a lot of situations where a Tesla car can, again, they can accelerate, can stay in the lane, can change lanes even, and can even park without the driver input. And that that's really a big advancement from 2013, where the INFINITI car was limited really to highway driving, but things like changing lanes and parking was not really within the wheelhouse of that vehicle back in 2013. Jumping forward a few years in October 2017, there was an auto truck, which is maybe not as well known, but it was purchased by Uber in 2019. And it delivered 50,000 bottles of beer after driving 120 highway miles in Colorado without a human at the wheel. And you can actually find a vehicle of this, they did a promotional vehicle. You can search for auto truck beer delivery, and you'll see the truck. The human will drive the truck onto the highway, but once it's on the highway, the human gets up, walks around the back of the cab, and the truck drives itself for 120 miles in Colorado, until it reaches its general destination. And then the human takes over again and brings the truck to its final destination for the last mile. This was considered Level Four autonomy because the technology was in control, but only active on the highway. So again, that's our definition. Level Four means that a human really does not need to take over, and that's the case here with the auto truck, the human got up and walked around, did not have to stay at the wheel, but it was only active on the highway. It's not Level Five because once we got to the final mile, once we got to city streets, the human had to take over again. So the human driver merged onto the highway and then the truck took over. And when it was time to exit the highway, the human got back into the driver's seat and controlled the vehicle for that last mile. The latest advance advancement here in the current state of technology is General Motors has an autonomous vehicle company called Cruise and that's actually operating a commercial driverless ride-hailing service in San Francisco right now. So if you travel to San Francisco, you can hail a Cruise autonomous vehicle. It's basically a taxi operating for commercial gain, and you can get in a car. There's no driver. You tell the car where to go, and it will take you there. And that's really one of the first examples in the US, at least of autonomous vehicles operating on a commercial basis, not just for testing, not just for limited research or testing instances. It's really an exciting advancement here that Cruise is operating their driverless vehicles for commercial profit, ride-hail services. You can go fly out there and hail one right now. So where's the technology going? Every automaker and many tech companies in Silicon Valley are racing to develop this technology. There's a lot of VC funding that's supporting this, a lot of mobility that's supporting this. A lot of states and other government entities are supporting this. And it's really an exciting time for these vehicles to be coming out and for people working in this industry. By 2024 General Motors promises that individuals will be able to buy their own self-driving cars. By 2030, Uber promises that its fleet will be completely driverless. So the ride-sharing service, they're saying in the next eight years, eight to 10 years, they're gonna be completely driverless. By 2040, Honda promises that there will be zero crashes in its vehicles by that year. And then the expectation is that by 2050, so looking out, 30 years from now, the expectation in the industry is that nearly all vehicles in US are able to drive themselves. However, these are, in some ways, optimistic predictions. There will certainly be challenges in this technology. The CEO of Waymo, which was previously Google's autonomous vehicle project, which was spun off into a company called Waymo. So he announced in November 2018 that autonomous technology won't be widespread for decades. And that autonomous vehicles will always be subject to constraints. And that was really a splash of cold water on the industry, to some extent, to have really an industry insider make these pronouncements. And even in the case of Tesla, which has probably, again, the most widely known Level Two vehicle operating in the US now, they had to walk back promises that its cars would reach Level Five by 2017 and that the company would have robotaxis on the road by 2020. They're obviously adjusting their predictions. So, it's not gonna be an easy transition to this driverless future. So the predictions that we mentioned in the previous minute or so, take it with a grain of salt, because there have been instances where these overly ambitious promises have had to be walked back as the technology is maybe not advanced as quickly as some have hoped. So what would this look like in practice? So before we have Level Four or Level Five autonomous cars before they become widespread on public roads, my expectation is that they're gonna be widely used, first in commercial settings. So we've already spoken about the example of the auto truck. I think a good first use case is highway driving for trucks with a human driver to drive the last mile. Highways generally are relatively stable speeds, no pedestrians, no animals running into the road in most cases. The roads are generally well maintained. Highway driving certainly is an easier driving task for computers, for autonomous vehicles, than city driving or sub-suburban driving, where a ball could jump in, could be thrown into a street and then a child runs out to get it. We're seeing now to find another example, we're seeing now that mining and other heavy equipment in sparsely populated areas are very popular for autonomous vehicles. There's a new story, a few years back, of a remote mine in Australia, far away from where everyone lived and it was hard to find drivers there. So that was a very easy and simple use case for these autonomous vehicles to be used to transport or transport other heavy mining equipment. when the humans were not there, didn't want to be there and it was a pretty simple use case. My expectation also is that we'll have what I call novelty settings, but I think they could be very important. Novelty settings, such as university campuses or theme parks. I kind of joke that if Disney World were being built today, we won't have a monorail. We would have an autonomous shuttle or some type of autonomous train that would handle the driving cast and get people from the parking lot to the theme park. I have a client that has spoken to a university and they basically said, we have all these college athletes that are practicing and exercising and working out all night and training all night, and it's about a mile from the training facility back to their dorm room. We don't want to pay a human driver to be driving the shuttle, to drive a mile 24 hours a day. They're looking essentially to have a autonomous shuttle that can drive this mile loop on a university campus. That's another example where it's a pretty close environment, low speeds, just a mile trip. And that would make a lot of sense for the type of initial use case where autonomous vehicles can make a big difference right now. Let's think for a moment about the potential benefits of widespread adoption of this technology. So it's important to start with the baseline of what the current state of mobility and transportation is in the US. Currently, with human drivers, and we're starting to see some advanced safety features, but still, it's mostly humans behind the wheel controlling vehicles. In the current state, we have about 10 million crashes in the United States each year. When I drove to my office yesterday, I saw two crashes, two separate crashes within about half a mile of each other. So we've all seen it. Unfortunately, it's very widespread. We're looking at about 10 million crashes in the United States each year. And that includes more than 38,000 traffic fatalities in 2020, which is the last year for which we have complete data. Those numbers have fallen down a bit, but in the last couple years, they've been creeping back up and a lot of people are attributing that to distracted driving, entertainment centers in cars, use of cell phones. We got down to a lower level in the 30,000 range, but it's crept back up in 2020, back up to around 38,000 traffic fatalities. Of the 10 million crashes that causes about 4.8 million injuries. And we're looking at about, 474 billion is a rough number, but that's the estimated cost of society for all these annual traffic crashes, fatalities, injuries. It comes to about $500 billion each year. And there's reason to think that autonomous vehicles can help. There's evidence that suggests that driver error causes 9.5 million out of the 10 million annual crashes. So about 95% of the crashes are caused by driver error, meaning it's not the fault of the vehicle itself. It's really the driver, either did something wrong, didn't see someone, was distracted. In 2020, the stats say that alcohol was involved in about 30% of all traffic-related deaths and the majority of drivers, ages 18 to 24, admit that they text while driving. So, the theory here is that if we can eliminate alcohol, eliminate distracted driving, eliminated texting, we can do a lot to drive down these crashes. And the theory is a computer doesn't get drunk, a computer doesn't get distracted. We can do a lot to bring down these crashes and really save a lot of lives and help make driving transportation a much safer activity. Beyond the safety potential, I think there are a lot of widespread benefits for this technology for autonomous vehicles where a human doesn't actually need to be behind the wheel in controlling the vehicle. First of all, this technology could allow for greater mobility for people who can't now drive. Either children, and we can talk about, maybe a public policy issue of how old you need to be to go in a driverless vehicle. I wouldn't wanna send a two-year-old in a driverless vehicle, but maybe a 15-year-old who doesn't have a driver's license. it wouldn't be a big deal for the driverless vehicle to take a 15-year-old to school or to other places. Because even if we don't trust the 15-year-old to control the vehicle, we trust them to be a passenger. People that are blind or have vision impairment or people that are elderly, or are not able to drive now, you can imagine lots of reasons. This could create a whole lot of mobility for people. And I've heard stories in the City of Trenton, the capital of New Jersey, there are a lot of residents who are excited about this technology, because right now it's hard to get around the city and they're looking for ways to increase their mobility and increase their flexibility and allow them to get around. So that's really one big benefit here is that if we can remove the driver from the equation, we can improve mobility for a lot of people. Another potential benefit is that we can transform countless hours of commuting time into productive time. If I don't have to be behind the wheel of the car and I could be on my laptop, that one-hour commute that I have every day or half hour commute that I have every day that could be productive working time. My car can become a mobile office. If there's nothing going on at work, I can then watch a movie or watch a TV show, or be on my phone. It could become a mobile entertainment center. So there's a potential here for commuting time to stop being wasted time and start being productive time again. Autonomous vehicles can also ease the transition to electric vehicles. All the clients that I work with now that are really planning five, 10 years down the road for what the future of transportation's gonna look like they're taking it as a given that their vehicles are gonna be electric, not internal combustion. So I think there's a potential here for these two technologies to dovetail the driverless and the electric vehicles can really work together to hasten the transition to electric vehicles. There's also an efficiency consideration here. Right now, my car sits unused, I'd say 90-95% of the time. I drive to work. And then my car sits in the garage for eight hours or 10 hours, or however long I stay at the office. I come home, and then the car sits around until I take my kids to practice or karate or whatever it is. But most of the time, my car just sits there doing nothing. It's not being used. Moreover, I have a five-seater car, and most of the time, I'd say, the four non-driver seats in my car sit empty. So about 90% of the time, when I'm driving to my office, when I'm doing my day-to-day activities, I have a five-seater car that's really transporting one person. That's a lot of wasted drive capacity. I obviously have a five-seater car because there are some cases where every seat is filled and I wanna have that flexibility, and I want to have a nice size trunk, and I wanna have a car that I like, and I'm not gonna trade in my car each time I need to go somewhere. But it's a lot of wasted efficiency and a wasted capacity where I'm not really using my vehicle to its capacity, 95% of the time. In the future, there's even thought that I may not even need to own a car. If an autonomous vehicle can be hailed as a taxi and drive me anywhere I want for little money, there's no need to pay a driver. I can really drive the 15 miles to my office. And it's cheaper than owning a car, cheaper than me paying insurance, cheaper than me maintaining the car and getting oil changes. Why would I wanna own a car if I can, for $5 a day, get my own ride into my office and go to work each day. So those are some of the benefits that are potentially coming with this new technology. And following up on that last point, I think there's really two potential models for car ownership in the future when this technology really matures and becomes ready for use. So Uber and other players in that realm, they believe that no one will own a car in the future for the reasons that I discussed. One estimate is that the cost for an autonomous taxi would be 35 cents per mile versus an average of about three and a half dollars per mile taxi cost in the United States. So if we can bring that cost down for a taxi down a 10th of what it was previously, it may make a lot more sense for me to pay, if I'm driving 35 cents per mile, and I'm driving 10 miles, that's only a couple bucks for me to get where I wanna go, get to my office and come back. I don't need to own a car if that's how cheap it's gonna be for me to get around. Other players like Tesla, believe in more of a mixed model. So their vision is that each family can own one vehicle, which can be shared among many family members. So the car can drive me to my office, can then come back and take my kids to school. Can then come and take my wife to her work. There's one car where, because it doesn't have to stay with me, the vehicle can go take me somewhere and then go back and then the rest of my family can use that vehicle. And then, if there's downtime during the day, if no one's using the car, then Tesla's planning on having a hailing fleet. A fleet of cars where my personal car can go and earn some money back for me. It can go act as a taxi that other people can hire, can rent for me, and I can go make back some of the price of the car by letting my personal car be used as a taxi during 90% of the day when it's not being used by me and my family. The other part of that vision is that Tesla would also operate its own fleet of taxis in cities where demand exceeds the supply of customer-owned cars. I think really the question here is, is society ready for autonomous vehicles. Waymo, which again is a Google spinoff, their self-driving vehicles have been involved in at least a dozen accidents. There was a fatal accident in Arizona in 2018 in an Uber vehicle. There was a fatal accident in Texas in 2021 in a Tesla. So again, the technology is not perfect now, and it's probably never gonna be perfect. So I don't think it's gonna be the case where there's never gonna be crashes. There's never gonna be accidents. There's never gonna be fatalities. Even if we totally remove the human from the driver, remove the human from the driver's seat. But it's important to keep in mind that the baseline is not zero accidents. There are already nearly 40,000 annual traffic deaths a year. And what can we do to bring those down? I think every life saved is precious. And if we can bring down those fatalities by even 5,000, 10,000 per year, that's really the promise of where autonomous vehicles can take us. Hopefully, we'll get it down to close to zero, but even if there's not zero fatalities ever in the future, I think we can do a lot better than 40,000 annual traffic deaths. However, I think it's really a question that I've had a lot of conversations with people I know, are you ready to drive in an autonomous vehicle? Are you ready to get into a car that you can't control? And I think that raises the question. Even if vehicles are statistically safer than human drivers today, if the computer's driving, are people going to accept the risk of giving control to a computer, right? The joke is that every person thinks that they're an above-average driver, and obviously, the math doesn't work on that. Not everyone can be above average, but people trust themselves, and they trust their family members to drive them safely. I've spoken to people in my family, they said I'm never gonna get into an autonomous vehicle. I don't trust the computer. I trust myself. As we know statistically, it's not necessarily a safe activity to drive a car, but people trust themselves, and they don't want to give up that control to a computer. However, this may be a generational change that we'll see take place in the next few years. It's important to note that norms can really change quickly. And that's true for societal norms also. Up until 1945, in New York City, all elevators were operated by attendants. No one will get into an elevator and press the button because that was not my job. I would get in, and then there would be a professional in each elevator in New York City that would control the elevator, and would make sure that we arrived safely at our destination. This is going up and down in a controlled environment, not on a public roads, but that was the norm until 1945. What happened in 1945 is that there was an elevator operator strike in New York City. And that spur of transition to having people control their own elevators, they realize that they can do it without an operator. And now we don't think twice. I don't mind at all getting into an elevator that doesn't have an operator. So, within the span of a few years, we really went from a situation where there was a professional, a human controlling an elevator to now, you just push a button and it takes you where you want to go. And that norm changed very quickly. Let's take a moment and consider the state of law as it is now. So who has jurisdiction over autonomous vehicles? Really, the answer is twofold. The laws are gonna overlap between federal and local jurisdiction. So the National Highway Traffic Safety Administration is responsible for a few things. Right now, they're responsible for setting and enforcing federal motor vehicle safety standards, which must be met before vehicles are sold. And it's also responsible for investigating and managing recalls and defects. The short answer is that the federal government, through the National Highway Traffic Safety Administration, they have jurisdiction over the vehicles themselves. The federal government's not gonna let you buy a vehicle that's not safe. That's not road roadworthy. And it makes a lot of sense that the federal government would be in charge of that because we don't want to have California and New Mexico have different standards or California and Texas have different standards and a vehicle that's legal in one state, should be legal in all states. So that's the role of federal government is to make sure that the vehicle itself is safe. On the other hand, states are responsible for licensing drivers, registering motor vehicles, enacting and enforcing traffic laws, conducting safety inspections, conducting emission inspections, and then also regulating insurance and liability. So the basic breakdown again, is that the federal government is in control of the vehicles themselves, but all parts of the driving task, all parts of the driving activity, whether that's insurance, whether that's rules of the road, traffic laws, licensing drivers, that's all the responsibility of the states. You don't go to the federal government to get your driver's license. You go to your state DMV. There's 35 states in Washington DC have actually passed laws related to autonomous vehicles. So it's pretty widespread. Most of the states now, majority of the states now has some type of law related to autonomous vehicles to try to prepare the legal environment for a future where this technology is widespread. In addition, the governors of Arizona, Delaware, Hawaii, Idaho, Illinois, Maine, Massachusetts, Minnesota, Ohio, Washington, and Wisconsin have issued executive orders related to autonomous vehicle. So even though there's not legislation in all those states, we've seen executive orders be promulgated in those states. What do those laws say? There's really a big range. Alabama's law basically does not too much. It simply establishes a joint legislative committee to study self-driving vehicles. So right now in Alabama, there's not a licensing regime. There's not really any type of permitting process. There's no insurance, special legislation related to insurance. What Alabama did was let's pass a law and let's go ahead and establish a legislative committee to study the issue, and we'll come back and do legislation when we've thoroughly studied the issue. In contrast, state like Arizona has one of the most permissive autonomous vehicle frameworks in the country. And that's mostly due to a series of executive orders signed by the governor. In Arizona, automakers need only to notify the Arizona Department of Transportation before they start testing. As long as their vehicles comply with state and federal laws governing motor vehicles. So as long as your vehicle basically meets federal safety standards and meets other local laws, state laws, You just need to notify the Arizona Department of Transportation, and then you can start testing in Arizona. You don't need to get a permit. You don't need to jump through a lot of hoops. You can basically just start operating there. You give them the notification, and then you're ready to go. This low barrier to entry has really made Arizona hotbed of autonomous vehicle innovation because they have a very welcoming regulatory structure. And it makes a lot of sense that companies have been eager to start testing and Arizona's a great place to do it. I think the weather also plays a role. You're not dealing with snow. You're not dealing with rain conditions, the same way that you would in the Northeast or other places. So Arizona's really been a hotbed of testing and innovation there. Waymo has been testing there for years. For example, the ride-sharing company, Lyft has also been testing there. And there have even been several Arizona cities that have launched ride-sharing autonomous vehicle programs on public roads within the last couple years. To take another approach, we see in California that the California Department of Motor Vehicles has three permit options for autonomous vehicles, for which a manufacturer can apply. So there's a testing permit that requires a driver. So you can test autonomous vehicles, but you still need to have a human in the vehicle to be able to take control and make sure that the vehicle's operating safely. And as of May 2022, 47 testing permits with driver have been issued as of a couple months ago. The second category in California is that you can get a driverless testing permit. So again, you can test in that scenario, you don't actually need a human in the car. You can't operate it for commercial purposes. You can't rent it out, but you can do testing without a driver. And again, as of May 2022 seven such permits have been issued in California. The last category here is what's called a deployment or public use permit. And that really allows you to operate as you would any vehicle. You can operate a taxi service. You can do commercial transportation. You can do commercial trucking. You can do all that without a driver and to deploy your technology, deploy your autonomous vehicles. And in California, as of a couple months ago, there are three such permits that have been issued. The law in California also requires that companies publicly report crashes and instances where human takes control of the vehicle. So you can actually go to the California DMV website and you can read incident reports that describe the time and date of an incident, the make and model of the vehicle involved, an assessment and description of vehicle damage, and then also narrative description. And again, that's true in cases where there is some type of crash or accident of the vehicle, or even a situation where the human feels the need to take over control, California wants to have a record of all that, and there's reporting that needs to be done. And it's kind of intriguing to go ahead and read some of those incident reports, but you can go ahead and do that. There's this big trove of data that California publishes each year based on all the testing that goes on there. In shifting gears a little bit, So in New Jersey, in March 2019, they established a... There was legislation that was enacted, and it had established an 11-member, New Jersey Advanced Autonomous Vehicle Task Force. And that task force was really made to conduct a study of advanced autonomous vehicles and make recommendations on laws and regulations that the state may adopt to safely integrate autonomous vehicles on the state roads. I'm really focusing on New Jersey for a few minutes because I was lucky enough to be appointed by the governor to serve on that task force. And it was a great experience to be working with people inside government, outside government, to think about the future of what autonomous vehicles would look like in the state of New Jersey, considering issues such as insurance, infrastructure. We had representatives from a car dealer, from an insurance company, engineering company, was really bringing a lot of people together to think through all issues, societal acceptance. What's the state gonna do with bus drivers or taxi drivers that find that may find themselves out of a job. And you can actually go read our, it's about a 70-80 page report that the New Jersey Autonomous Vehicle Task Force published. And again, so I'll focus on New Jersey just because I'm proud of my experience serving on that task force. One of the innovations that we had on that task force is really focused on licensing and permitting for testing and use of highly automated vehicles on public roadways. And some of the recommendations that we made was first of all, there should be a permitting process to allow companies to test and deploy autonomous vehicles on public roadways, which, by the way, is not such an obvious thing, but that was our recommendation. I've heard a call. I received a call from a potential client a couple months ago, and they said we operate in a bunch of states. We have a customer in New Jersey that wants to hire us to do transportation. It wouldn't be passenger. It'd be commercial use, basically to help companies move goods from one place to another, in the logistics field. And this company called me up and said, we gotta call from a client in New Jersey. They want us to come operate our trucks, our driverless trucks in New Jersey. And we can't figure out if it's illegal or not. And I told them, that's a very good question. It's not clear if it's illegal or not. There's nothing that says, yes, it's legal, And here's how you get a permit, or here's how you get a license to operate or test your autonomous vehicles on New Jersey roadways. But there's also not anything that says you can't do it, right? There's no law that says you can't operate autonomous vehicles in New Jersey. The vehicle safety laws and the traffic laws in New Jersey they're silent on the issue. They just basically assume that there's gonna be human driver behind the wheel. So I had this conversation with this potential client. And they said you know what, we're not gonna be operating in New Jersey. We wanna make sure that we go through the permitting process that we get a license. If there's a gray area, we're just gonna stay out. And they gave up, they essentially gave up a customer because the law was unclear here in New Jersey. My expectation is that when there is a permitting process or there's a licensing process, they're gonna be one of the first ones to go ahead and apply for that license because they wanna do business here. And the ambiguity in our laws is really keeping them from operating in the state. One other thing that we worked on in our task force is that we made a recommendation to the legislature that there should be reciprocity in permitting and licensing in New Jersey and other states. So we identified a number of states that had sufficiently rigorous licensing processes. So a state like Michigan or California, they do a good job vetting their vehicles and vetting their users. So if you have a license in California or Michigan, or a few other states, the recommendation that our task force made to the state was essentially we should have accelerated process, and we should basically allow those companies to operate in New Jersey, because they were already jumping through the hoops, and they've already proven that they can operate in other states, so let's make it easy for them to come here to our state. But beyond this one piece, this one bill, or this one piece of legislation that established the task force, New Jersey has not seen other legislation be enacted that's focused on autonomous vehicles. There have been a few bills that have been introduced, and I think it's helpful to consider them for just a second, even though they're still, have not become law because the bills that have been introduced in New Jersey have been very similar to what we find in other states that except for an outlier like California or Arizona that we've spoken about before, most states that have legislation related to autonomous vehicles, follow these same types of general concepts. So one example would be that a vehicle needs to be operated by a person designated by the automaker. So that essentially means that there needs to be a safety driver in the vehicle, that you can't just let it operate without a human. That's different than California, which does have a process for allowing people to drive without a human. But this bill and which is very common across the states, is that there still needs to be a safety driver. And more than that, the safety driver needs to be sitting in the driver's seat the entire time and needs to be capable of immediately taking control of the vehicle. Another aspect is that there is heightened insurance requirements. So it's very common. You see the number of $5 million or $10 million. Mostly the two common numbers that you see, but many of the places where there's legislation, there's heightened insurance requirements, either a minimum of 5 million or minimum of $10 million, that would be the insurance minimum, to operate or test in the state. So moving on from the state level, and we've given just a brief overview of the legislation involved in different states, there's obviously more that we didn't cover, but moving on to the federal level. There were two bills that were under consideration by Congress addressing autonomous vehicles. One was called the AV START Act, and one was called the SELF-DRIVE Act. So the SELF DRIVE Act actually passed the House of Representatives in 2017, but was not approved or not passed by the Senate. And so neither bill advanced the passage, neither the AV START Act nor the SELF DRIVE Act advanced the passage, and we're still waiting for federal legislation on this piece. The SELF DRIVE Act was reintroduced again by Congress in 2021, but again has yet to be passed. And the federal government's foot-dragging here has created a bit of a backlog for people that wanna operate or test autonomous vehicles because under current federal law, the National Highway Traffic Safety Administration needs to grant exemptions for autonomous vehicles that don't meet traditional vehicle standards. And the catch is that they're limited to only 2,500 exemptions annually. And you can imagine that those exemptions get snatched up quickly. It's not enough to keep up with demand for vehicles that don't meet the safety standards. And you think of things like side view mirrors, is that gonna really be important, or other safety features that are relevant for human drivers? Is that gonna be important when a computer drives, the vehicle controls the vehicle? Probably not, but the federal law still mandates safety features that are attuned to human drivers. And right now, the only way around that is to get one of these limited exemptions. It's not a sustainable model. So where's the law going? So back in 2016, the National Highway Traffic Safety Administration issued non-binding guidelines for the testing and deployment of autonomous vehicles. There was a 15-point safety checklist for autonomous vehicles that covered things like data privacy, post-crash behavior, object detection, cybersecurity, human and machine interface, and crashworthiness. Again, remember that we discussed before that the National Highway Traffic Safety Administration has the power to identify safety defects and to recall vehicles. So, if you want to operate autonomous vehicles in the United States, you really need to satisfy that agency to make sure that they're not going to recall your vehicle or allow it on the road in the first place. You really need to be consistent with their guidelines for things that we may not have thought relevant to vehicles before. But things like data privacy and cybersecurity, I think those are gonna be increasingly important issues when it comes to transportation. The National Highway Traffic Safety Administration also published a model state policy that hopefully would be adopted. The hope was that it would be adopted by all the states or many of the states. And the goal is to avoid a patchwork of different laws that are in place in different states. And when it comes to transportation, that's especially important because I want my car to be able to be driven from one state to another. I wanna be able to drive my car from North Carolina to South Carolina. That's the whole point of a car is to go from one place to another. And if I can't drive my vehicle across state lines, that's gonna create a real challenge for people. Jumping up a few years in January 2020, the National Highway Traffic Safety Administration issued a report, a white paper called, Ensuring American Leadership in Automated Vehicle Technologies: Automated Vehicles 4.0. That builds upon previous versions of this report by really looking at 30 relevant US government components, which have direct or tangential equities in safe development and integration of the AV technologies, and really trying to get a whole of government approach to focus on the safe deployment, the safe development of these technologies to really focus on not just on highway safety, but other areas where the US government's involved to try to move things along. In just a few months ago, in March 2022, the National Highway Traffic Safety Administration, again, enacted a final rule, establishing federal safety standards for autonomous vehicles. So we were talking about before the federal laws basically have certain safety standards that are really geared towards human drivers and not technology, not computers. So there was a final agency rule that came out in March 2022, that really was focused on updating safety requirements for occupants in vehicles that don't have traditional manual controls associated with a human driver. So part of this is a change in terminology that would reflect the spatial layout of automated vehicles and build on really the agency's previous guidance to ensure public safety as automation evolves. And it focus on things like the term driver's seat or steering wheel or passenger seat. Those spatial references don't really make sense for purpose-built autonomous vehicles that aren't built with such features, right? I don't need a driver's seat. I don't need a steering wheel. So it's important to update that terminology and update those standards. And we're starting to see movement for that from the federal government. But really, again, if I'm no longer driving a car, you can rethink what the entire vehicle looks like, right? There no longer needs to be a steering wheel. There no longer needs to be driver seats. It could be one bench that everyone... Think of a shuttle that you can come in, you can plug in your laptop, you can have a desk because really I'm not going to, and I'm not expected to be able to control the vehicle. So we've discussed a bit about the current state of the law, where the law may be going in terms of regulation, but there's still open questions under the law. And the big open question under the law is even if this technology dramatically reduces crashes and fatalities, and damages, who's gonna be liable for the accidents that do occur. And related to that question is what's that gonna mean for the insurance industry? What's the insurance regime gonna look like? In my mind, ultimately, the question of who's gonna pay for accidents is easy, and who's gonna pay for damages is easy. At the end of the day, the customers are gonna pay for accidents one way or another, but how we get there is really anything but easy. So let take a step back and consider where we are today. So today, most states follow traditional auto liability. That means that for minor crashes, car crashes, there's gonna be a claim adjuster at insurance companies that are gonna allocate blame based on conventions, right? So the driver that rear-ended another car is gonna be at fault. If I'm driving on the highway and someone rear-ends me, the person behind me is gonna be at fault and their insurance company's gonna have to pay. Again, those are for minor crashes. For more major crashes, there's gonna be litigation, settlements. That's gonna really determine liability by closely analyzing negligent standards. And that's really where we are today with liability under the current format, which made a lot of sense when humans are driving the vehicle. In contrast to that, there are 15 states that take a no-fault approach. And the no-fault system basically means that the legal system is not gonna be involved at all, unless the statutory threshold for seriousness is met, right? So there are different standards for that. If the crash or the accident was serious enough in one of various ways, that's when the legal system gets involved, but in all other cases, and for those serious crashes, those major crashes, that's when, again, you can go to court, you can sue, you can enter settlements, you can analyze negligence and liability. But except for those major cases, there's not even an effort to allocate fault. And instead, each insurance company simply pays its own insured for the damage suffered. So we're not gonna worry if you rear end me, your insurance company is not gonna have to pay the brunt of that. Instead, each insurance company really takes care of its own drivers, its own policyholders. So we're not gonna even take the time to allocate fault or determine who is responsible, who is liable. You just take care of you, your people, I'll take care of my people and that's the end of the story. And the only way that that goes out the window and that we do actually try to find fault or determine liabilities is when certain thresholds of seriousness are met for a crash. So that's kind of where we are now with liability. There have been some suggestions for where we may go with this in the future. And I think it's a good question. There are lots of law review articles. There are lots of different approaches here. We're gonna consider a few in the remaining time that we have. So there's some law review articles that suggest that we should borrow a model from federal legislation that would preempt tort liability for vaccine manufacturers. And this is not to get into any current controversies, but back several decades ago, there was concern that there would be tort liability, too much tort liability for vaccines for various diseases. And the federal government wanted to encourage vaccine uptake and use of vaccines. And so there was legislation that preempted any type of traditional tort liability for vaccine manufacturers, right? So if I was injured by a vaccine, I couldn't sue the manufacturer, but instead, what Congress did was they created a special forum that would dispense with all claims that anyone would have against the vaccine manufacturers. So the model here would be, let's take that same approach. Congress could create a special forum to deal with all claims that this would be funded by contributions from the automakers and the manufacturers. So instead of me suing a vehicle that injures me or damages me because I was not controlling the vehicle, the vehicle was driving by itself and it crashed into something and I hurt my shoulder, so I'm not gonna be able to then sue the manufacturer. I would have to take my claim to the special forum. The automakers would put money into that. And this would really provide cost certainty to automakers and allow for efficient disposal of claims. It's just a more streamlined system. And it would really encourage people to develop autonomous vehicles if they know that they're not gonna have to deal with one-off claims, there's gonna be a special forum to deal with all this. The downside of this approach is that it's gonna deny fair compensation to injured parties, or potentially, at least in some cases, if there's a special forum, there may be limitations on damages. You don't get your day in court, if you have to go through the special forum. They're only gonna put in certain amounts of money. So that's really the downside of this approach, but that's been... Some law reviews articles have suggested this special forum approach to preempt liability because we wanna encourage the transition to autonomous vehicles, so let's make it easier for the manufacturers to do that. Another approach here would be to institute strict liability on automakers. And the theory is if humans are not in control of the car, they should not be liable for any damage that the automaker's engineers caused, right? I can't control it. There's no steering wheel, what do you want me to do? I bought the car, or I rented it, or I hailed it as a taxi, but I can't control where the vehicle goes. I can't move the vehicle if I see someone running to the street or there's another car that sways in front of me, why should I be liable? It's the automakers and their engineers who didn't do a good enough job avoiding damages. Even if that's the case, even if there's gonna be strict liability, those costs will ultimately be paid for by consumers, right? Automakers can build the cost of damages into the purchase price, right? So if I purchase a vehicle that drives itself and I can't control it, and there's no steering wheel, maybe I'll pay extra dollars to handle liability. That's gonna be worked into the purchase price of the car. When it comes to robotaxis or ride-hailing, maybe you add 10 cents per ride that would help cover liability costs and damages costs because the automakers are ultimately gonna do that and they can spread the cost. Ultimately, I think why this is appealing to me at least, or at least not as an initial step in this process is that the manufacturers are in the best position to understand and spread the risk, right? They can understand, all right, our vehicles, we know that it's gonna be better than 40,000 or 38,000 fatalities each year. We know there may be some we're gonna have to pay out. In these situations, we may have to pay out damages. They're gonna be in a position to understand that risk, spread the risk, and price it appropriately so that those costs would ultimately again be paid for by consumers. But it gets paid upfront, or it gets paid, the compensation is paid for by consumers upfront, or when they hail a taxi. This approach is also somewhat appealing because it provides very concrete incentives to automakers, to improve safety. If you don't wanna pay out, under the strict liability regime, then improve your cars and improve safety and make sure that they don't get into crashes. And in fact, Mercedes and Volvo have already voluntarily accepted liability, in part because of these reasons. To take a step back a little bit and let's consider the three objectives of driving, under our framework of what are the open questions under the law. So there's really three objectives of driving if you think about it. There's a mobility objective, I wanna get to get from place to place. There's a safety objective, I wanna make sure everyone gets there safely. And there's a legality objective. I wanna do that in accordance with traffic laws. So people may have heard about the trolley problem, right? If I can't stop a crash, this is like a philosophical, ethical question, moral ethics. If I can't stop a crash, can you divert the trolley to run over other people to save yourself? Can you divert the trolley to, if there's three people standing on one spur and one person standing on another, who do you save if you can't save everyone. So the same thing is true for autonomous vehicles. There may be some situations in which ideal operation of the vehicle is not possible. So in a simple case with our three objectives, what should an autonomous vehicle do when a truck is blocking a single lane? Imagine a single-lane highway a truck is broke down, is blocking the vehicle. So something needs to be compromised. So my autonomous vehicle can wait and not drive anywhere, but that compromises my mobility objective. I'm not getting where I want to go. It can proceed that compromises its safety objective, right? It can just drive straight ahead, but that's not gonna be safe. Or the vehicle can also cross the yellow line, right? If it's a one lane in each direction, the vehicle can cross the yellow line that would achieve mobility. It would achieve safety, but it would compromise the legality objective, right? That's not in obeyance with traffic laws. And why this is interesting is that, if I'm stuck doing that, if I'm driving in the car and that scenario happens to me, I can control the vehicle. I can guess what I would do. Everyone can make a decision on their own what they would do, but when it comes to autonomous vehicles, again, if there's no steering wheel, there's no way for me to control the vehicle, then those questions and those considerations need to be thought of by the engineers and the philosophers and the ethicists who are sitting right now in Michigan or Silicon Valley or wherever else, they need to be thinking about these issues now and figure out how to deal with that situation. The much harder cases arise when a crash is unavoidable, right? So, in the case that we spoke about before, you can easily avoid a crash. You may be delayed, you may compromise the traffic laws, but ultimately, everyone's gonna be safe, hopefully. The harder question is when a vehicle's outta control, something needs to get hit. There is gonna be a crash in some way, it's unavoidable, so what should the car do? And how should we prioritize the safety of people in the scenario? And again, this is not, the difference is this is not a snap decision that you have a second to react or a split second to react. And you don't really have time to think through the implications. Right now, we have plenty of time to think through the implications, and we should have people start thinking about them. So in that scenario where a crash is not avoidable, should the priority be to save the most number of lives? Studies respond and say that, yes, that should be the priorities, to save the most number of lives, but it gets a lot more thorny. Should the car prioritize, and when I say the car, I mean, the engineers and the ethicist working on these issues, should the priority be to save a baby over a person who's a hundred and three years old. Mercedes has come out and announced that its cars are gonna be programmed to prioritize the safety of the passengers in the car, right? It's not gonna prioritize the safety of pedestrians or other cars. Really, if you get into Mercedes, you know that the priority's gonna be protecting the safety of people in the vehicle. People have made fun of Mercedes for that, or tease Mercedes for that saying, yeah, of course, if you buy a Mercedes, then you're gonna protect yourself. But I think if we look at that more broadly, would you buy a car that is not gonna have a priority of protecting you in the car, right? Is the car gonna sacrifice you as the customer, to save someone else? Is that gonna slow and hinder the widespread adoption of this technology if people understandably wanna be a little bit selfish for that, in that way? Is that gonna slow the widespread adoption? Maybe it's best if every car is programmed to protect itself and that's gonna be the standard. People can make fun of Mercedes for that, but I think it's a bit more complicated than it may appear at first. And the last part of this is should these decisions be made by automakers, and that can conclude engineers or their ethicists or their philosophers? In my mind, this really should be a public policy decision, either decided by the legislative process, the regulatory process. I think it's a conversation that we as a society need to have, figure out how to prioritize saving people in these scenarios. Hopefully, they'll be few and far between, but how do you prioritize the safety of individuals? Should we leave that up to the manufacturers and the automakers and trust that they'll do their best, or should this really be a matter of public policy? So to conclude, I wanna just spend a few minutes thinking about what a self-driving future will mean and what it will look like. So, we spoke about the federal state breakdown of regulation, but I think it's also important not to overlook the role that cities are gonna have to play in this future. There's one study done a couple years ago of 41 major cities, and they found that an average of 31% of urban space in downtown areas is devoted to parking. So if I have a meeting in downtown Manhattan or Midtown Manhattan, and my car can park itself, can drop me off at my meeting and then park itself in the suburbs or in a different borough, and then come get me when my meeting's over, that really eliminates the need for parking in our urban environments, in our cities, how will dense cities be transformed when the needs of parking are much alleviated because I may not even own a car. I get in a self-driving taxi. It drops me off. I don't have to worry about parking at that point. And then it goes to move... Then it moves on the vehicle and picks up a new passenger. So I think it's important to start thinking about how our urban environments can look and what they will look like when parking is either reduced or removed from the equation. What role will cities have in determining whether to allow autonomous vehicles, right? If a state says that it's okay and there's a permitting process, is there gonna be a role for cities to opt-out of allowing those vehicles in your city? And I think the opposite approach may also be relevant. If and when autonomous vehicles are safer than human drivers, will there be any states, or will there be any cities that allow only autonomous vehicles within the jurisdiction? Are we gonna get to a point where only autonomous vehicles they're so safe that we don't trust humans. We don't allow humans to drive their vehicles on public roads. If you wanna go drive your car, you take it to a racetrack. It's kind of like horse horseback riding. We don't use horses to get around too much these days, but people still riding horses. So you go out to a farm, you go out to a track, you ride your horse for fun. It's a leisure activity. You can envision the future in cars as well, where I don't need to... I'm not allowed to drive my car on public roads. It's not safe enough. The computers much safer than me. The autonomous vehicles are much safer. So if I wanna drive my car, I'll have to do it elsewhere on a designated track, sign liability waivers. And that's the only time where humans can drive the car. And our last piece of business here. So I just wanna think for a moment about what self-driving future, what a self-driving future will mean for different entities. So I think car dealers should start thinking about this. Auto dealers and car dealers under an Uber model that we discussed earlier where people don't buy cars, they just get into a cheap taxi everywhere they wanna go. Is there still gonna be a need to buy cars? How should the car dealers react to that? The same is true for insurance companies. Will individuals still need to purchase insurance like we do now to operate my own vehicle? Or is it gonna be the case where if the manufacturer's liable and I'm not really controlling the vehicle, will the insurance companies need to start advertising to the car manufacturers and not to individuals, right, to handle their insurance needs. Will those manufacturers self-insure. I think there's a lot of ambiguity and a lot of unknown and uncertainty in what insurance is gonna look like. The same is true for real estate. We spoke about this briefly before, but if 31% of downtown areas is no longer need for parking, what's that gonna mean for real estate in our cities, if people accept longer commutes, if driving time becomes productive time or working time or entertainment time, I don't mind getting in my car and driving for two hours because I can be on my laptop. That's working time. I don't need to live close to my office. I don't need to live close to people because I can just... I don't mind wasting that time anymore driving long distances. I think the same question needs to be asked about our driving infrastructure. What are our roads gonna look like when they no longer need to be suited for human drivers? And to some extent, good roads are good for anyone for, humans or computers, but exit signs, lane markings, what's that gonna look like? What's our infrastructure gonna look like when humans are no longer driving? And the last thing to consider here is what that's gonna mean for taxi and truck drivers, right? Right now, there are approximately 200,000 taxi drivers and 3.5 million truck drivers in the United States. How will society handle their gradual or potentially sudden unemployment? And I think we should be thinking about those issues now, so that there's smooth transition in all these factors for car dealers, insurance companies, real estate infrastructure, employment matters. I think there's a lot of things beyond just the technology here that we need to be considering as a society. Thank you.