Benn’s latest video is good. However the rhetorical conceit that cars are equivalent to people is not productive. I’m commenting on the video because it is a strong and substantial argument, versus so many drivers online merely complaining about their driving behavior being tracked.
The “you” in the title, and the message itself in the video https://youtu.be/uB0gr7Fh6lY&t=8m47s, erases people who do not use cars for transportation or at least do not identify themselves primarily as drivers. Which in America largely means people who are financially or physically unable to drive. Genuine and valid criticisms of Flock shouldn’t rely on language that erases those groups.
Furthermore, this framing is not helpful to people who do rely on cars for transportation. There is a lot of troublesome surveillance. Some people identify strongly as drivers. A proper understanding of surveillance and privacy for them includes distinguishing threats tied to driving versus other threats they are exposed to. Also, conceptualizing people as drivers implicitly cuts off consideration of the most important way to avoid automotive surveillance, i.e., finding lifestyles that don’t depend on driving for people who can manage to do so.
The video largely ignores the genuine harms that drivers cause which warrant surveillance. That’s not a both-sides statement because it’s not a defense of Flock: these Flock cameras are obviously not optimized for traffic safety. The point is we have a century of data, even before these tech companies started selling these cameras, showing that rising use of cars in public means more interaction between individuals and law enforcement. The visceral understanding that something has to be done about driving enables all sorts of things to be done, albeit largely in bad faith. So merely getting rid of ALPRs is not a sufficient proposal if there’s a goal of privacy beyond avoiding pictures of cars. Genuine opposition to surveillance of cars from an anti-surveillance perspective should elevate the need to shift away from cars to other forms of transportation which are more amenable to privacy from the government. In addition to mode shifting, it’s important to reduce the potential for surveillance of other forms of transportation, e.g., advocacy for free fares and resisting abuse of cameras aimed at people in public.
The video mentions that Flock takes pictures without a license plate, and it needs to do that to check if there's a plate. Even if there's no plate there may well be a car. There's no way around the need for review the image more holistically to get at the nominal goal of tracking vehicles, and it’s especially important to track vehicles that have no visible license plate. It's odd that images stay on the device as long as the video indicates, possibly without being transmitted, but the location of the information doesn't seem like the biggest problem so much as who gets access to it. The video also mentions that hypothetically Flocks could be pointed at a front door, but so could non-ALPR cameras. It’s reasonable to worry about video surveillance of front doors, it’s a much broader and largely unrelated issue to ALRPs.
The video refers to both the security of the technology itself and privacy. Both are valid but they are also different concerns. Flock and associated companies should be more proficient about securing their own devices. The video does a great job of demonstrating their lack of technical competence. A hypothetical hacker could access some portion of the data. Some larger unauthorized entity could potentially access all of their data. That privacy violation would be bad in itself, and could lead to scams, so enforcing greater security is good. But if Flock etc were perfectly secured and only authorized users could access the data, the privacy risk would still be almost as large as it is now. Those unauthorized users could still physically install their own cameras. Meanwhile privacy would still be vulnerable to many law enforcement personnel authorized to use the data, and they have the power of the government at their disposal. There should be rules about how that data is used in terms of privacy. Principally I’d like DOTs to monitor their motor vehicle users instead of transit agencies tracking their riders, without making it a law enforcement matter.
Effectively reducing crime or catching dangerous offenders depends on use. My sense is that Flock is used for hit and run crashes often enough to be a net positive, but that reflects the general abysmal state of traffic safety such that anything helps. Flock and other ALRPs seem to be set up intentionally to minimize the number of drivers caught committing driving offenses. I once asked for Flock data to try to identify a drunk driver who was passing near children, and seeing the pictures made it clear the camera had been positioned to not capture much other than the rear of a vehicle and a license plate. But most drivers concerned about the anonymity of their cars would be livid if efficacy were increased by detecting more drivers speeding, looking at phones, running reds, driving through crosswalks while pedestrians have right of way, etc.
There’s a real tension between improving safety on the roads and that aversion to surveillance of cars, and not just among Benn and people aligned with his way of thinking. Consider the people who like Flock as it exists. Let’s say people who are stereotypically tough on stereotypical types of crime reported by police and news papers, who are strong supporters of car dependency as a way to keep non-drivers away from them, and who don’t view traffic offenses as real crimes. Those Flock supporters do not want to receive tickets themselves for violating those laws. HOAs paying for Flock cameras don’t want to get complaints from residents if Flock were actually designed to capture dangerous behavior by drivers in the neighborhood. My interest is primarily in safety, although I understand other people have different priorities. But I would suggest people in the former group (prioritizing opposing surveillance of cars) should consider why they in practice land so close to the latter group (surveillance of cars, but only for locations) and merely disagree about location tracking. The best way to reduce the safety risk of cars, and to reduce surveillance of cars, is to have fewer cars.
The video makes a strong case for a licensing regime for vendors selling these sorts of technologies to the government https://youtu.be/uB0gr7Fh6lY&t=38m44s . However Benn says "you can't legally drive a car past a Flock safety camera without taking a routine drivers test and having a vehicle that passes basic safety requirements" as an argument, and we should aim higher. In GA, where I believe he lives, drivers can get a license as children and never be tested again. The roads are full of objectively unsafe vehicles. There is nominally a licensing and testing regime but it's an example of what to avoid, not a model to replicate.
Benn mentions a Flock camera he cannot avoid very close to his driveway in a rural area. It seems like a surprising allocation of resources to have a Flock camera on that road. The whole country has 80,000 cameras, that’s a lot but not enough for every neighborhood. My main suggestion, as mentioned above, to shift aware from cars is hardest to implement in those neighborhoods. So his dissatisfaction with the situation is understandable. That being said, traffic deaths are a particularly high risk for rural America, presumably including hit and runs although I don’t have statistics handy. As stated above I agree with him about Flock’s security problems. But even if these were fully publicly owned ALPRs with well tested software this would still presumably feel weird to him. I don’t know that there’s a feel good option. Letting drivers keep killing anonymously is not acceptable, at least to me. ALPRs seem to be the best way for now to detect cars (although especially in rural areas, the problem seems to be trucks) driven by some of the most dangerous drivers.
For example, I am writing this on Nov 16, minutes after this story was published https://www.wsbtv.com/news/local/airman-killed-hit-and-run-while-changing-tire-georgia-interstate/LU4UZHICMZFAJHVCUUJQ7PL6RA/ although it’s about a crash on Nov 2. I didn’t even search for this for the purposes of this post; due to my particular interest in drivers killing people this was on my regular news feed. Apparently a driver used a truck with a trailer to kill this man changing his tire on the side of a road in rural Georgia. Then the driver fled the scene. Two weeks have passed and there’s no suspect, i.e., that driver is still driving. Potentially still driving in rural Georgia, although the crash was close to the state line. The driver hit this victim on a highway, not a residential street. But it’s going to be hard to get anywhere much without going on highways, so if ALPRs were only on highways I don’t see how that significantly changes the privacy risk.