• Tesla launches ‘actually smart summon’ or ‘ass’ (Electrek, September 3, 2024)

  Tesla has introduced the long-awaited Actually Smart Summon (ASS) feature, alongside a simpler Dumb Summon option, in its latest software update. Actually Smart Summon enables Tesla owners to call their vehicle from a parking spot to a chosen nearby location via the Tesla App. The release notes humorously describe ASS as a magical tech experience, allowing your car to autonomously navigate to your position or destination.

  Dumb Summon offers basic functionality, allowing the car to move forward or backward with app-based controls. To use these features, owners simply access the Summon tab in the Tesla app, gaining live camera feeds to oversee the vehicle’s movements. While employing these features, vigilant supervision is advised, particularly in parking scenarios. Currently, ASS is available for cars equipped with Hardware 4, with future availability for Hardware 3 suggested imminently.

• Woman using hands-free driving feature hits car, killing 2 in Pennsylvania, cops say (Miami Herald, September 3, 2024)

  A Pennsylvania woman, Dimple Patel, turned herself in on Tuesday this week after hitting and killing two men while driving her Ford Mustang Mach-E in the vehicle’s hands-free driving mode back in March of this year. At around 3 AM on March 3 in Philadelphia, two drivers were pulled over on Interstate 95 when Patel’s yellow Mach-E slammed into a vehicle from behind, killing both men. The Mach-E is reported to have been traveling at around 70 miles per hour with  BlueCruise hands-free driving feature enabled, as well as Adaptive Cruise Control when the crash occurred. Ford has commented on the incident, stating, “There is no priority higher than safety at Ford, and we have been collaborating fully with all relevant organizations on the investigations into these incidents.”

  The marketing materials for Ford’s BlueCruise state that drivers should continue to keep their eyes on the road when BlueCruise is enabled. Pennsylvania State Police urge drivers to “be aware and ready to regain control if the situation warrants your immediate reaction.”

• Researchers probe safety of AI in driverless cars, find vulnerabilities (Tech Xplore, September 2, 2024)

  Artificial intelligence (AI) is a critical component in self-driving vehicles, facilitating decision-making, sensing, and predictive modeling. However, its vulnerability to attacks is a concern being explored by researchers at the University at Buffalo. Their studies indicate that malicious actors could potentially manipulate these AI systems, causing them to fail. For instance, 3D-printed objects might camouflage a vehicle from AI radar systems, rendering it undetectable.

  Although current autonomous vehicles remain safe, this research could impact the automotive and tech industries, and influence regulatory policies. Tests conducted at UB have shown that the most commonly used radar systems for object detection, though reliable, can be compromised using strategically placed “tile masks.” These discoveries underline potential attack motives such as insurance fraud or competition, and highlight the gap in security measures, which researchers aim to address by developing robust defenses.

• Tesla reports profit drop on price cuts, lower vehicle sales (France 24, July 24, 2024)

  Tesla reported a significant second-quarter drop in profits after aggressive price cuts on vehicles and steep expenses for reorganization of the company after laying off 10% of its staff earlier this year. The carmaker reported profits of $1.5 billion in the second quarter of 2024, a 45% drop from last year. Despite the dramatic drop in sales, Tesla feels optimistic about its plans to begin production of its more affordable Tesla models in early 2025. The company also expects its Cybertruck vehicle to stay on track for profitability by the end of 2024.

  As we recently reported, Tesla has also postponed its robotaxi event that was scheduled for August, which “allowed [Tesla] to improve the robotaxi as well as add in a couple of other things for the product unveil,” stated Elon Musk after moving the event to October.

  It seems that despite a poor Q2, the automaker remains “overly optimistic” about its future and the likelihood of fully autonomous robotaxis, Musk adding that he “would be shocked if we cannot do it next year.”

• GM indefinitely delays Cruise Origin autonomous vehicle (CNBC, July 23, 2024)

  General Motors has indefinitely delayed production of its Origin autonomous vehicle as it navigates difficulties with its self-driving unit, Cruise. Instead, the focus will shift to using the next-generation Chevrolet Bolt for developing autonomous technology. This decision, outlined by GM CEO Mary Barra, is aimed at reducing costs and mitigating regulatory issues linked to the vehicle’s lack of manual controls. The production halt follows a November incident where a Cruise robotaxi in San Francisco dragged a pedestrian, prompting a suspension of driverless operations and a $600 million setback in GM’s second-quarter earnings. Former Cruise CEO Kyle Vogt expressed disappointment, noting GM’s history of losing technological leads. A third-party review found that regulatory oversights contributed to the San Francisco accident. Cruise, with major investors such as Honda, Microsoft, and Walmart, is now resuming supervised driving in other cities but has not yet relaunched in San Francisco.

• British motorists overwhelmingly reject self-driving and AI vehicles despite road safety benefits (GB News, July 19, 2024)

  A recent survey conducted in Brittan shows that the public is not ready for self-driving vehicles on the road.  Only 37% of those polled support autonomous vehicle technology.  Not surprising, adults between the ages of 18-44 were more in favor of the technology vs those over 45.  When asked what bad habits could be cut out by self-driving vehicles, being impatient, distracted and exceeding the speed limit topped the list of answers. 

  Simon Staton, client management director for Venson Automotive Solutions said, “While many company car drivers may shy away from switching to autonomous as soon as they become legal, acceptance will grow rapidly, and many will be keen to explore the options.”

• Generative AI could power the next wave of self-driving cars (Axios, June 18, 2024)

  The future of self-driving cars may rely on advancements being made in artificial intelligence in order to accelerate the development of AVs in a more capital-efficient manner. Many believe that AV development is stuck, as many automakers struggle to advance the technology needed for self-driving vehicles while maintaining profits. Many investors are calling the next phase of autonomous vehicle development, “AV 2.0.” Companies, such as Nvidia and Waabi, are creating self-driving systems for vehicles using AI as a tool to make “human” decisions safely and correctly from intuitive learning.

  “Learning-based AVs can be taught driving skills more quickly, but the black-box nature of machine learning-based behavior makes it more difficult to validate safety,” states Phil Koopman, associate professor at Carnegie Mellon University and AV safety expert. “Even if you drive a billion miles in simulation, you still need to make sure the simulation wasn’t missing edge cases that will cause crashes in the real world.” The world of AI is still far from perfect, but as the technology develops, the stakes for autonomous driving will become much higher.

• China tops the world in driverless car testing (MSN, June 15, 2024)

  China has embraced driverless car technology like no other country in the world.  In Wuhan, tech giant Bidau has 500 driverless taxis on the road, many of which operate without safety drivers, and plan to add another 1,000 to its fleet. Currently, sixteen cities in China participate in driverless vehicle testing with at least nineteen automakers and their suppliers vying to establish dominance.  The Chinese government plays a large role in the success of the autonomous vehicle sector as it controls the narrative about safety, designate areas for on-road testing, limit discussion with online censorship, and have stringent data control policies. Countries like the US and Japan continue to test driverless car technology, but at a more cautious rate.

• GM pumps $850 million in Cruise to keep struggling robotaxi company afloat (The Verge, June 11, 2024)

  General Motors is injecting $850 million into Cruise to support its operations following the highly publicized suspension of its robotaxi service, after a driverless car was involved in a pedestrian accident. The funding will help Cruise resume testing in various U.S. cities and seek additional investors. Despite significant losses — $8.2 billion since 2017, including $3.48 billion in 2023 — GM remains committed to Cruise. After the incident, Cruise recalled all 950 vehicles for a software update, launched a safety review, and faced executive resignations. Co-founders Kyle Vogt and Dan Kan, plus a quarter of the staff, departed, leading to leadership changes, with GM executives taking key roles. Cruise is cautiously returning to the roads with safety drivers in cities like Houston, Phoenix, and Dallas, and continues its efforts towards fully autonomous operations.

As widely publicized, numerous automakers and technology companies have sunk exorbitant amounts of money into the race to commercially scale self-driving vehicles.  Although there has been significant progress over the past few decades, the final stage of the journey toward truly hands-free driving is proving to be quite the challenge, with many paths being pursued to achieve the ultimate goal of full autonomy.

Despite the rapidly changing landscape as new technologies emerge, the race to autonomous driving has essentially become a two-horse contest between companies focused on the gradual evolution of technology stacks from Level 2 autonomy to full self-driving and companies focused on immediate Level 4 and 5 technology stacks.

As to the first category of competitors, OEM’s such as Tesla and technology companies such as Mobileye are currently leading the way in pursuing Level 2 “plus” autonomous driving systems with the goal of gradually integrating more advanced autonomous capabilities into consumer vehicles.  By offering beta-testing of driver assistance systems on a progressive basis for an upfront fee or subscription, these companies have uniquely been able to create revenue streams that enable them to subsidize the colossal costs associated with developing commercially viable “fully” autonomous driving systems.  This financial bridge allows for the transition of autonomous content and continued rollout of next level applications without worrying about burn rates and lack of new funding sources.   

On the other end of the spectrum are companies who are attempting to develop Level 4 and 5 vehicles or “robotaxis” from scratch through billion-dollar capital investments in autonomous technology, such as Cruise, Waymo, AutoX, DiDi, Pony.ai, Zoox, Aurora, Motional, and Optimus Ride (acquired by Magna). Although there was great exuberance that this perpetually unprofitable space was close to achieving commercial viability, a growing body of real-world data has recently provided a more sobering assessment.

In the aftermath of the widely publicized Oct. 2 crash in San Francisco where a driverless car ran over a pedestrian, Cruise suspended all U.S. operations, announced significant cutbacks in resource allocations and saw top executives resign.  Additionally, there have been numerous reports of robotaxis blocking streets, failing to properly recognize emergency vehicles or emergency scenes, or simply freezing indecisively in intersections or roundabouts.

Although solving complex urban driving scenarios in a scalable way has long been the holy grail for the AV industry, these recent controversies have inevitably raised questions as to whether current driverless technology is ready for wide scale deployment in major American cities.  In fact, Argo AI, which had received billions in funding from Ford and Volkswagen, shut down and completely abandoned its autonomy pursuit earlier this year.  Despite these headwinds, the remaining robotaxi companies and mobility providers are publicly staying committed to realizing a future of safe autonomous mobility, even if may take more time than initially anticipated.

So, who will win the race to commercially viable self-driving?  The answer seems to lie in what sector will be able to generate sustainable revenue while the technology continues to develop and emerge.  Despite promises that the dream of driverless vehicles is just around the corner, the reality is that it will likely be at least 10 years before the technology is ready for full-scale deployment.  According to Jeremy Carlson, Associate Director for the autonomy practice at S&P Global Mobility, “Level 5 Autonomy, which refers to a vehicle that can mimic human driving abilities in any environment, is unlikely to be available to the public before 2035 and perhaps even later.”  Most industry experts agree that autonomous vehicle technology will be limited for the foreseeable future to geofenced robotaxis operated by fleets in predefined zones and hands-off systems with driver monitoring devices in personal vehicles that will still necessitate driver involvement.

The real challenge is that vehicle autonomy cannot be scaled until current safety margins or MTBF (Mean-Time-Between Failure) rates, which measure the average time a system can operate without a failure, surpass that of human drivers.  Existing technology stacks simply have not reached this safety threshold and additional progress in this regard is proving to be sticker than expected.  At present, Level 4 and 5 companies are still facing billion-dollar investment needs to accomplish this goal.  Given astronomical burn rates and a potential 10-year runway to achieve success, the future does not look extremely positive for any company that has not figured out how to offset development costs with revenue streams.

Thus, it appears that companies like Tesla and Mobileye, who are transitioning autonomous content gradually with built-in revenue streams from existing ADAS, clearly have a leg up on the rest of the field.  In fact, the only viable solution to sustainability challenges may be to build existing markets for partial autonomy that can fund the roll-out of more sophisticated systems on a revenue-driven timetable.

As more and more industry experts are acknowledging that the race to full autonomy is a marathon rather than a sprint, automakers and technology companies are beginning to shift resources to the business models pioneered by the likes of Tesla and Mobileye.  The inescapable reality appears to be that reaching optimal levels of vehicle autonomy will require substantially more time and exorbitant additional investment.  Unfortunately, these are not luxuries that non-revenue generating Level 4 and 5 ventures can afford, at least not as currently structured.

Given the life altering benefits vehicle autonomy will eventually provide and the enormous profits that will be realized by first movers in the industry, we have no doubt that the vehicle autonomy wars will continue to be interesting and heated.  Although the safe bet currently appears to be companies systematically transitioning from partial to full autonomy, the fascinating part of this race is that technological breakthroughs can quickly change race standings at any given moment.

• Tesla wins bellwether trial over Autopilot car crash (Reuters, April 22, 2023)

  In 2020, Justine Hsu brought a personal injury lawsuit against Tesla in a California state court, claiming that her Model S swerved into a curb while operating in Autopilot, causing the airbag to deploy “so violently it fractured [her] jaw, knocked out teeth, and caused nerve damage to her face.” On Friday, April 21, following a three-week trial, a jury returned a complete defense verdict for Tesla, finding that the “Autopilot feature did not fail in what appeared to be the first trial related to a crash involving the partially automated driving software.” Hsu was awarded none of the $3M in damages that she sought in the suit. Her attorney, Don Slavik (who has brought multiple actions against Tesla), “expressed disappointment.”

  Tesla’s primary defense in the suit was that Hsu did not follow the user manual warnings against using Autopilot on city streets. Jurors apparently agreed that these warnings were sufficient, telling Reuters that “Tesla clearly warned that the partially automated driving software was not a self-piloted system, and that driver distraction was to blame.”

  Georgetown Law professor Ed Walters called the verdict a “huge win” for Tesla, adding that “[t]his case should be a wakeup call to Tesla owners: they can’t over-rely on Autopilot, and they really need to be ready to take control and Tesla is not a self-driving system.”

  The case docket can be accessed here (Case No. 20STCV18473). Additional reporting on the verdict from slightly different angles is available from Insurance Journal, Automotive News, Inside EVs, Fortune, and Marquette University Law School’s Faculty Blog.

• California lawmakers looking to ban automated trucks with no safety driver (Overdrive, April 24, 2023)

  After passing out of the California Assembly Transportation Committee, California Assembly Bill 316 recently withstood another test when it passed out of the state Committee on Communications and Conveyance. 

  As we have noted previously, the legislation, authored by California Assembly Members Cecilia Aguiar-Curry (D-Winters), Tom Lackey (R-Palmdale), and Ash Kalra (D-San Jose), would prohibit the operation of an autonomous vehicle with a gross vehicle weight of 10,000 pounds or more on public roads for testing purposes, transporting goods, or transporting passengers without a human safety operator physically inside the vehicle at the time of operation.

  In an attempt to protect human trucking jobs, the Teamsters and the California Labor Federation have lobbied heavily in favor of the legislation. Prior to the hearing, Randy Cammack, President of Teamsters Joint Council 42, commented that “[t]he public should not be treated as a lab rat for big corporations to test their technology.”

  On the other side of the aisle, more than 55 business organizations and companies signed a letter opposing the bill. Per the letter: “AB 316 continues to be an effective ban on this life-saving technology with no pathway for Californians to benefit from autonomous trucks. … Supporters of AB 316 have shown zero evidence that AV trucks are unsafe or will cause mass layoffs of truck drivers. This committee is being asked to take action based on nothing more than speculation.”

  Notwithstanding the eventual outcome of the pending legislation, the California Department of Motor Vehicles is considering a regulatory framework that would allow for autonomous vehicles on California roadways without consent from the state legislature.

• Ford Establishes New Division For AI & Autonomy (Clean Technica, April 20, 2023)

  About a month ago, we featured an article about Ford’s new AI subsidiary for automated driving technology, Latitude AI. This more recent article is notable (and worth a read) for its in-depth look at the history of autonomous vehicle developments leading up to Ford’s work, going back to the 1939 World’s Fair where GM sponsored a “Futurama exhibit showing off carts controlled by transmitters placed in roadways.” Early V2I? Other historical harbingers of autonomous technology include experiments with pavement-embedded wires and radar-enabled “Roadarscope” in the 1950s and 1960s, and DARPA’s Autonomous Land driven Vehicle project in the 1980s and 1990s.

• GM to end production of electric Chevy Bolt, its first mass-market EV, later this year (CNBC, April 25, 2023)

  GM announced yesterday that it will stop production of its original and best-selling long-range EV, the Chevy Bolt. The Bolt also is currently the cheapest EV on the market. During an investor call, CEO Mary Barra explained that GM will retool the Bolt plant for production of electric trucks. A Chevy spokesperson added, “[w]hen the Chevrolet Bolt EV launched, it was a huge technical achievement and the first affordable EV, which set in motion GM’s all-electric future. … Chevrolet will launch several new EVs later this year based on the Ultium platform in key segments, including the Silverado EV, Blazer EV and Equinox EV.”

OpenAI’s chatbot, ChatGPT, which debuted in November, has quickly become a viral sensation especially in the field of content creation. GPT-4, the language model that powers the latest version of ChatGPT, has been popping up more and more frequently as the real author behind content generated for myriad wide-ranging audiences.  But is AI really ready to render content writers obsolete by generating high-quality content at the click of a button?

To explore this question, we decided to jump on the AI bandwagon and asked ChatGPT to contribute to the blog this week.  Specifically, we asked ChatGPT to “write an article on the dangers of autonomous driving vehicles.”  In response, ChatGPT generated the below content in less than 10 seconds:

Autonomous driving vehicles have been touted as the future of transportation, promising to make our roads safer, reduce traffic congestion, and improve overall efficiency. However, despite their potential benefits, autonomous driving vehicles also come with their fair share of risks and dangers.

One of the most significant dangers of autonomous driving vehicles is the potential for software malfunctions or errors. Autonomous driving vehicles rely on complex software systems to operate, and any malfunction or error could lead to catastrophic consequences. For example, if a self-driving car fails to detect a pedestrian or misinterprets a traffic signal, it could result in a serious accident.

Another danger of autonomous driving vehicles is their susceptibility to hacking and cyber attacks. Since these vehicles are connected to the internet, they are vulnerable to cybercriminals who could exploit any weaknesses in their systems to take control of the vehicle or compromise its data. A cyber attack on an autonomous driving vehicle could lead to a loss of control or a crash, putting passengers and other road users at risk.

Moreover, autonomous driving vehicles could create new ethical dilemmas that we are not yet equipped to handle. For example, if a self-driving car is faced with a situation where it must choose between two potentially fatal outcomes, such as hitting a pedestrian or crashing the car, how should it decide? These kinds of ethical dilemmas are difficult for humans to navigate, let alone machines.

In addition, there is the issue of legal liability. In the event of an accident involving an autonomous driving vehicle, who is responsible? Is it the vehicle manufacturer, the software developer, the vehicle owner, or the passenger? As the law currently stands, it is unclear who bears the legal responsibility in such cases, which could lead to lengthy and costly legal battles.

Finally, there is the danger that autonomous driving vehicles could exacerbate existing social inequalities. Since autonomous driving technology is expensive, it is likely that it will only be accessible to the wealthy, further dividing society into those who can afford the latest technology and those who cannot. Furthermore, autonomous driving vehicles could also lead to job losses in industries such as transportation, which could disproportionately affect low-income workers.

In conclusion, while autonomous driving vehicles hold great promise for the future of transportation, they also come with significant risks and dangers. As we continue to develop and refine this technology, we must also consider and address these potential dangers to ensure that autonomous driving vehicles are as safe and beneficial as possible for all members of society.

Although the above content was accurate, it also was very basic—in other words, it hit the top-level points of the AV debate, but it definitely did not capture any nuance.  Our experiment with GPT-4 revealed that it didn’t really add much substantive value or new insight; rather, it simply regurgitated ideas and observations that have already circulated extensively on the internet.  Because of AI’s current inability to genuinely simulate true human insight, interpretation, and analysis, our verdict is AI cannot outperform human content creators. At least not yet.

What do you think?

Copyright Nelson Niehaus LLC

The opinions expressed in this blog are those of the author(s) and do not necessarily reflect the views of the Firm, its clients, or any of its or their respective affiliates. This blog post is for general information purposes and is not intended to be and should not be taken as legal advice.