top of page
Writer's pictureEvan Atlas

In Defense of Spot

I recently read a New York Times article on the NYPD’s use of Spot, the robot dog built by Boston Dynamics. It’s a subject I’ve given a bit of thought in the last few years. It’s part of a broad acceleration towards an economy in which robots, artificial intelligence, brain-computer interfaces, and the Internet of Things (IoT) will be just as important as human intelligence. I also see this specific application of robotics as an opportunity to synthesize some the world’s disparate views on justice, police accountability, and systemic racism.



The NYT article highlights the “dystopian” aesthetic of a robot police force with a quote from City Councilman Ben Kallos, who draws a comparison to an episode of Black Mirror. There is a robotic dog in the episode which is very Spot-like, except that it acts completely independently, and is portrayed solely as a human-hunting weapon. Beyond that, there is little to say. The article could have easily invoked RoboCop, but its cultural ubiquity already makes it a touchstone in any conversation about police and robotics. Chappie is too cute, and cybernetic Matt Damon too much of a man-of-the-people, to evoke the kind of terror needed to nudge people towards the view that NYPD has Terminator-envy. But this sensationalized view of robotics limits our ability to make real progress.


The article also quotes Councilman Kallos saying that “At a time where we should be having more beat cops on the street, building relationships with residents, they’re actually headed in another direction in trying to replace them with robots.” But the truth is that a statement like this is, at best, myopic. The integration of robotics and policing is, in fact, one of the few policies that should be as appealing to those who emphasize that either Black, Blue, or All lives matter.


To imagine that robo-pups are primarily going to replace cops patrolling the streets is missing the main selling point of Spot. The NYT quotes a representative from Boston Dynamics who says that “out of the roughly 500 robotic dogs that are in the field worldwide, most are being used by utility companies, on construction sites or in other commercial settings that involve dangerous situations.” Nobody, to my knowledge, has called Spot “dystopian” in any of these other contexts.


Let’s see if we can get past gut-reactions and movie-induced anxieties, and examine why this idea is such a win-win.


The NYT article echoes commonly expressed sentiments that police departments are too “militarized”, that police robots will inevitably be weaponized and used to kill or hunt down civilians, and that it represents another step towards Orwellian surveillance and the death of privacy. Let’s take these on one at a time.


While there is no arguing that police have quite literally been militarized, in that they are equipped with surplus military gear, this refrain seems to ignore that US citizens can purchase almost all of the same gear and weapons. In the wave of 2020 protests against police shootings, many protestors and counter-protestors were more well-armed than the police. If you accept the premise that we need some kind of police, or, more broadly a system of justice, then it should be obvious that the law can only be upheld through a relative monopoly on violence. Whether we are talking about the military of a whole nation or the police force of a small town, the asymmetry of power between law-enforcers and law-abiders is necessary if we want to embody our deepest values. If we don’t have a system of collective value-enforcement, then we cede value-generation to those who are most willing to use war as a means to that end. The age of 3D-printed guns and smart bullets and drone swarms is here, and will be unforgiving to the unprepared.


The second point, that we wouldn’t be able to keep guns out of the paws of robot dogs, is false and an unfair framing of the real issue. This fear glazes over a glaringly-obvious question: Why would your standard-issue Spot need a weapon? Quite the contrary, one of the advantages to sending robots to do jobs that are highly dangerous for humans is that they would not need weapons like their squishy human colleagues. We should be asking instead how the use of robots like Spot could be inversely-correlated with both harm to civilians from police, and harm to police from civilians.


It is anything but a foregone conclusion that robots would need to be weaponized in order to have this effect of harm reduction across the spectrum – from criminal to police officer to bystander. And beyond that lack of underlying need, nothing is preventing us from regulating the use of police robots, from the local level up to international efforts to ban lethal autonomous weapons.


Consider any scenario in which police officers must engage with people in their communities, and notice the areas of harm reduction that rarely get centered in this conversation. Police centaurs could, on the law enforcement side, dramatically reduce the frequency of life-threatening, adrenaline-spiking, PTSD-inducing incidents in this line of work. An officer who makes it to retirement without firing a gun or being under fire is nonetheless primed to die young.


A national shift in perspective has made commonplace the idea that we are expecting police to do too many different jobs, while receiving fewer hours of training than a barber. A holistic view of the situation is that 911 emergencies are not easily dichotomized between police department and fire department. I think it will seem natural for some combination of police officers, medics, mental health professionals, firemen, and robots to respond to emergencies. Or to have highly-trained “guardians” with skills in all of these areas. Having Spot in the mix would make all of these jobs safer and less traumatic for the humans involved. And if “hurt people hurt people”, then we should recognize any harm reduction as a victory for all, whether the primary change is on the side of police or civilians.



So what about privacy? Is our future full of robot dogs and insect-sized surveillance drones? To really get at the heart of this, we need to consider data rights more broadly. The current state of data privacy is abysmal. And that is in no way specific to this conversation. There is movement in the right direction with ideas like a data bill of rights, or the California Consumer Privacy Act, but a bigger wave is building. Spot needs a friend called “blockchain”.


We need to extend the question, in this case, beyond “Who owns your data?” What if Spot was not a police dog, but rather a sovereign dog – a liaison between the police and community? Spot, or other robots, could be owned and managed by a Decentralized Autonomous Organization – a blockchain-powered group of stakeholders in your community. This should satisfy advocates on opposite sides of the political chasm. Spot would be helpful to police, but accountable to those being policed. In theory, maybe a community could vote through their DAO to attach a laser cannon to Spot’s head, but would they? With so much pushback against police use of force, it seems dubious that a community would choose this path for themselves.


A DAO-operated robot would not only guarantee data privacy by storing sensitive data on the blockchain, it would actually reduce the need for other kinds of mass surveillance which put your personal data in the hands of governments and police by default. Monitoring isn’t our problem, per se – it is simply one of several necessities for effective governance of commons. Unaccountable surveillance in the hands of centralized authority is another story. But with decentralized organizations oriented towards the wellbeing of communities, Spot will be watching, but we will be watching the watchers. If a court can grant a search warrant to police, then, in that same spirit, a DAO could grant access to Spot’s hard drive. If, for example, Spot captures video footage of a school shooter, it seems likely the stakeholders of that Spot’s DAO would vote to release that data for the purpose of justice.


Beyond the knee-jerk fears of RoboCop and Terminator, as well as the legitimate but surmountable concerns explored above, a safer and more just world awaits us. Spot has the potential to be more like man’s best friend than the nightmarish “Metalhead” from Black Mirror. But we do need the right balance of monitoring and privacy. We need the right rules in place, as with any technology. And we need a model of organization and management, like the kind offered by DAOs, which puts power in the hands of the people most affected by this potential future. Taken all together, Spot would not represent something being “done to us” by external authorities, but rather a reflection of what values we wish to collectively uphold.


I’d like to offer, here, a Metalhead-alternative: the 2018 film A-X-L. Stories are important. Representation is important. And A-X-L, despite being a rather silly and predictable movie, adds nuance that Metalhead doesn’t. [SPOILERS follow]


The movie’s eponymous robotic dog, A-X-L, gets its name from its intended uses: attack, exploration, and logistics. It was developed by the military, and meant to assist humans on the battlefield. But instead of seeing it go to war, it escapes its human creators and winds up befriending a teenage boy named Miles – giving the film an aesthetic close to Transformers or E.T.


Most of the film consists of Miles and his love-interest, Sara, keeping A-X-L out of trouble and out of the hands of the military. And doing sweet dirt-bike tricks… At one point A-X-L turns himself (I’m using the pronoun Miles uses, since he objected to calling A-X-L “it”) into a ramp. Yes, really.



But what’s most salient about this movie is the tug-of-war between violence and peace, between designed purpose and practical use. In the hands of the military, A-X-L would be deadly, and dare-I-say, dystopian. But in the care of Miles, the robot is much more like the faithful canines he resembles. However strange this may seem, the movie stands out among the media mentioned earlier as the most faithful to reality. It is not a simple utopia or dystopia. The world is very much like ours, and points us towards the most mature way to see robots like Spot: neither doom-harbinger or panacea. The most natural comparison, then, is not to Metalhead or Terminator, but to a real-world technology like CRISPR. Opportunity and danger come hand-in-hand. But, if given the choice, I think many people would be happy to have Spot join their community.



Comentarios


bottom of page