Public Safety

The Future of AI in Self-Driving Vehicles

Kevin Boyle, MJLST Staffer

 

Last week, artificial intelligence (AI) made a big splash in the news after Russian President Vladimir Putin and billionaire tech giant Elon Musk both commented on the subject. Putin stated that whoever becomes the leader in artificial intelligence (AI) will become “the ruler of the world.” Elon Musk followed up Putin’s comments by declaring that competition for AI superiority between nations will most likely be the cause of World War III. These bold predictions grabbed the headlines; but in the same week, Lyft announced a new partnership with a company that produces AI for self-driving cars and the House passed the SELF DRIVE Act. The Lyft deal and the House bill are positive signs for investors of the autonomous vehicle industries; however, the legal landscape remains uncertain. As Putin and Musk have predicted, AI is certain to have a major impact on our future, but current legal hurdles exist before AI’s applications in self-driving vehicles can reach its potential.

 

One of the legal hurdles that currently exists is the varying laws between state and federal authorities. For example, Companies such as Google and Ford would like to introduce cars with no pedals or steering wheels that are operated entirely by AI. However, most states still require that a human driver be able to take “active physical control” of the car to manually override the autonomous vehicle. This requires a steering wheel and brakes, which would make those cars illegal to operate. At the federal level, the FAA requires that commercial drones be flown by certified operators, not computers or AI. Requiring operators instead of AI to steer drones for deliveries severely limits the potential of this innovative technology. Furthermore, international treaties, including the Geneva Convention, need to be addressed before we see fully autonomous cars.

 

The bipartisan SELF DRIVE Act recently passed by the House attempts to address most of the issues created by the patchwork of local, state, and federal regulations so that AI in self-driving cars can reach its potential. The House bill proposed clear guidelines for car manufacture guidelines, clarified the role of the NHTSA in regulating automated driving systems, and detailed cybersecurity requirements for automated vehicles. The Senate, however, is drafting its own bill for the SELF DRIVE Act. This week, the Senate Commerce, Science, and Transportation Committee will convene a hearing on automated safety technology in self-driving vehicles and the potential impacts on the economy. The committee will hear testimony from car manufacturers, public interest groups, and labor unions. Some of these groups will inevitably lobby against this bill and self-driving technology for fear of the potentially devastating impact on jobs in some industries. But ideally, the Senate bill will stick to the fundamentals from the House bill, which focuses on prioritizing safety, strengthening cybersecurity, and promoting the continued innovation of AI in autonomous vehicles.

 

Several legal obstacles still exist that are preventing the implementation of AI in automated vehicles. Congress’ SELF DRIVE Act has the potential to be a step in the right direction. The Senate needs to maintain the basic elements of the bill passed in the House to help advance the use of the innovative AI technology in self-driving cars. Unlike Musk, Mark Zuckerberg has taken a stance similar to those in the auto industry, and believes AI will bring about countless “improvements in the quality of our lives,” especially in the application of AI in self-driving vehicles.

 

 


Permissionless Innovation or Precautionary Principle: The Policy Menu of the Future

Ethan Konschuh, MJLST Staffer

In their recent paper, Guns, Limbs, and Toys: What Future for 3D Printing?, published in the Minnesota Journal of Law, Science, and Technology Volume 17, Issue 2, Adam Thierer and Adam Marcus discussed the potential regulatory frameworks for technological innovations that could spur what they call “the next great industrial revolution.”  They believe that 3D printing, one such innovation, could offer such great benefits that it could significantly enhance global welfare.  However, they worry that preemptive regulations on the technology could undermine these benefits before giving them a chance to be realized.  The paper advocates for a method of regulation called “permissionless innovation,” as opposed to regulations following the “precautionary principle.”  While there are many pros to the former, it could leave unchecked the risks curtailed by the latter.

“Permissionless innovation refers to the notion that experimentation with new technologies and business models should generally be permitted by default.”  It follows from the idea that unless a compelling case can be made that a new invention will bring serious harm to society, innovation should be allowed to continue unabated, and problems, should they arise, can be addressed later.  The authors point to numerous benefits of this approach with respect to emerging technologies.  One of the most obvious benefits is that this type of regulatory framework does not prematurely inhibit potential benefits.  “Regulatory systems based on precautionary thinking focus on preemptive remedies that aim to predict the future and its hypothetical problems. But if public policy is rooted in fear of hypothetical worst-case scenarios, it means that best-case scenarios will never come about.”  It would also preserve the modern startup culture where “just about anyone can afford to launch a business.”  Implementing a framework based on the precautionary principle will create barriers to entry and raise the cost of innovation.  This would also reduce the ability to maximize competitive advantage through trial and error, which refines the technology and efficient allocation of resources for development.  As an example of the potential detriments to competitive advantage from preemptive regulation, the authors point to the different policies of the Europe and the U.S. in the mid-nineties internet explosion where the former preemptively regulated and the latter allowed for permissionless innovation, resulting in the U.S. being a global leader in information technologies and Europe lagging far behind.

An alternative regulatory approach discussed in the article is based on the precautionary principle, which generally refers to the belief that new innovations should be curtailed or disallowed until it can be proven that they will not cause harm.  This approach, while posing problems of its own discussed above, would solve some of the problems arising under permissionless innovation.  While there are many economic and social benefits to permissionless innovation as the bedrock on which policy rests, it inherently allows for the “error” half of “trial and error.”  The whole concept is rooted in the idea of ex post regulation, creating policy to correct for problems that have already occurred.  While traditionally, as shown through the internet regulation difference and outcome between Europe and the U.S., the risk of error has not outweighed the benefits that result, new technologies pose new risks.

For example, in the realm of 3D printing, one of the hot topics is 3D printed firearms.  Current laws would not make 3D printed guns illegal, as most regulations focus on the sale and distribution of firearms, not creation for personal use.  The reasons why it might be more prudent to adopt a precautionary principle approach to regulating this technology are obvious.  To adopt an ex post approach to something that could have such dire consequences could be disastrous, especially considering the amount of time required to adopt policy and implement regulations.  Permissionless innovation could thus become a sort of self-fulfilling prophecy in that major tragedies resulting from 3D printing could result in exactly what advocates of permissionless innovation seek to prevent in the first place: strict regulation that undermines the development of the technology.

The debate will likely heat up as technology continues to develop.  In the era of self-driving cars, private drones, big data, and other technologies that continue to change the way that humans interact with the world around them, 3D printing is not the only area in which this discussion will arise.  The policy decisions that will be made in the next few years will have far reaching consequences that are difficult to predict.  Do the economic and social benefits of being able to manufacture goods at home outweigh the risks of legal, discrete self-armament and its consequences?  The proverbial pill may be too large for some to swallow.


A View to the Development of the FAA’s Ban on the Use of Galaxy 7 Phones on Airplanes.

Joshua Wold, MJLST Staffer

Samsung’s Galaxy Note 7 has problems. Under normal charging and use conditions, the battery in some devices can start on fire. As of September 1, Samsung reported that 35 of these problems had come to its attention, and more have been reported since that time. Samsung has already begun a recall—officially a “replacement program”—offering to replace the potentially dangerous devices with new ones. At the same time, US government agencies are also moving to prevent the harm that a malfunction from these devices could cause.

The Federal Aviation Administration (FAA) has not banned the device from airplanes. According to Matt Novak at Gizmodo, the FAA stated on September 6th that it was “working on guidance related to this issue,” and said, “If the device is recalled by the manufacturer, airline crew and passengers will not be able to bring recalled batteries or electronics that contain recalled batteries in the cabin of an aircraft, or in carry-on and checked baggage.” On September 8th, the FAA issued a statement “strongly” advising airline passengers to keep the devices off, not charge them while aboard, and to keep them out of checked baggage.

Then, on September 9th the Consumer Product Safety Commission (CPSC) put out a press release urging Galaxy Note 7 owners to power off the device, and not to charge or use it. The press release indicated that it was cooperating with Samsung “to formally announce an official recall of the devices, as soon as possible.”

Based on the comments made to Gizmodo, the FAA appeared unwilling to ban the device until the official recall was put into effect. This is understandable. Cell phones are a significant part of modern life. Nearly every person getting onto a plane is carrying at least one. TSA would thus be forced to distinguish between types of cell phones. Its task would be complicated by cases and other types of personalization which obscure the appearance of a cell phone.

Even more challenging, however, is the fact that some versions of the Note 7 have a battery which is not prone to overheating, and poses no threat. Unless these safe phones were to also be banned, security personnel would need to determine which battery was in a phone in order to know if it were permissible. People want to have their phones when they get wherever they are flying to, and banning a safe phone because it looks like an unsafe phone seems like a sure recipe for passenger dissatisfaction.

It may seem that a ban is be appropriate despite the difficulties. Certainly, the potential for harm is significant. A widely circulated photo of a Jeep engulfed in flames is evocative of the threat, and airline passengers have gotten used to restrictions on items which seem to pose even less risk than an exploding phone. FAA’s suggestion that those with Note 7 phones simply turn them off and not charge them may have had the potential to eliminate the threat. On the one hand, it was only a strong recommendation, and not a rule. On the other hand, with airlines repeating the FAA’s warning, it seems unlikely that many people would have failed to take it seriously.

On the 15th of September, the situation changed further, as the CPSC announced an official recall. With that decision, the ground mentioned to Gizmodo for not instituting a ban disappeared. Considering  some data (which can be found here), suggesting that people aren’t really taking Samsung’s warnings seriously, it seemed very likely that the FAA would decide to strengthen their recommendation against use to a prohibition on use, or even a prohibition on flying with the phones at all.

On the next day, September 16th, the FAA banned use of the phone on airplanes, but not the phones themselves. This new policy fits with the reality of modern cellphone use, that people rely on their phones, even if they are fire hazards. While this move takes the pressure off the TSA (which is probably a good move in terms of the overall happiness of air travelers), the regulation (which can be found here) doesn’t specifically mention the Galaxy Note 7, but refers instead to “defective or recalled” lithium batteries.

Of course, this creates the same sort of enforcement problems as appeared with earlier recommendations: can airline staff identify a Galaxy Note 7 with such “defective or recalled” lithium batteries? The FAA itself notes that it is difficult to distinguish between phones which have had the battery replaced and those which are still risked. The FAA’s recommendation on how to manage this problem is pretty general, and it essentially boils down to training of airline staff, and provision of information to airline passengers. One hopes that this is sufficient.