police technology

In the future, the police as we know it today will not exist.

This is not a political statement, it’s simply a technological fact. Now, it’s essential to remember that all technological facts are endlessly contingent. For example, it’s a technological fact that if you click on a link, another webpage will open. But that’s contingent, usually on some very complicated and impressive infrastructure operating without fault (or rather, with sufficient fault-tolerance whose few exceptions did not affect the expected outcome, this time).

If you click a link, it will only do what technologists expect if you’re using a browser that doesn’t have the wrong kind of malicious software. And you have to be using a computing device that doesn’t have some other hardware or software flaw that will prevent expected actions. And you have to be connected to a network that has sufficient range and capacity. And then an entirely different set of computing devices needs to be connected and operating as expected. And then all of that has to work correctly, walking backwards, in high heels. During this entire time, every device involved needs to have electric power in the right amount and at the right time. That is a lot of contingencies.

But still: if you click on a link, a webpage will open (even if it’s not the one you expected). And with just as much certainty: in the future, the job of police will not exist as we understand it today. That is a technological fact, and it requires very little understanding of technology to see that. It merely requires obvious extrapolation from technologies you see around you every day.

Most people, including most police officers, may think the job of police is to stop crime. But all police officers know that it’s more an exception than a rule that they make an arrest on any given day. This is not an indictment or a criticism in any way, it is simply a pure accounting of time. Cops probably spend 50% of any given day in travel time, going from place to place. Maybe another 20% of the day is talking to people: talking to each other, to dispatchers, to citizens with a question or complaint, to witnesses, to victims, to prosecutors and lawyers and judges and juries. Then 30% of the day is administrative: paperwork, paperwork, paperwork, court time, occasionally some training. As a proportion of time spent, there is almost no time spent on a usual day in the active act of stopping crime. Stopping crime might be the reason for police, but that’s not how they spend their time on the job.

Of course, there are occasions where crime is discovered during travel time noted above, and during the talking time above. That happens a lot more on TV than in real life. More often, crime is discovered through other means: an alarm, a call to 911, while carrying out a search warrant, perhaps during a stakeout, or a successful search for a suspect. Police action in each one of these cases is planned beforehand, it doesn’t happen extemporaneously. There is forewarning, and police are specifically sent to a location where the crime may be discovered. None of this is the result of random discovery during the usual day at work.

Technologists hate inefficiency, and can’t help but think about designing for a more efficient police force. A perfect police force would do nothing but fight crime: they would only conduct the very few activities that are a result of planned actions expecting to find crime. The other activities would be done by people who were not police: all that traveling around, talking to people, filling out paperwork – people who are not police officers can do all those things. That is not to diminish the importance of any of those things, and many of them are essential to stopping crime – they are just not themselves the active act of stopping crime that requires the most prepared police action.

In a perfect world, anyone who might ever be involved in actively stopping crime would spend all their free time preparing for the most dangerous police actions, and they would have exactly the resources they need to stop the most deadly opposition that they are likely to encounter – no more and no less. Because some crimes are so inherently dangerous, perfect police would spend all their time on training when they weren’t actively in the act of stopping crime. And in a perfect world, their training would be perfect, so they would follow the best possible tactics to avoid escalation and the use of deadly force, including the elimination of any kind of bias whatsoever.

Obviously, we do not live in a perfect world. There are many, many social reasons why we cannot today operate a perfect police force. Many. But there are also many technological reasons: we cannot predict where crime will happen, we can’t be everywhere at once, we can’t travel fast enough or efficiently enough or safely enough. We might not have the data we need to identify everything that we need in order to make good use of technology, including data relevant to both crime and to training.

The thing about technology is, though, that all of the technological problems will be solved, so long as social barriers don’t prevent that from happening. To be clear: this is NOT an argument for the moral supremacy of technology. Morality is only to be found in society, not in technology – and there may be times when the development of a certain technology may be itself immoral. However, in the absence of social barriers (including moral barriers that we should respect), technology problems will be solved, because that’s the definition of technology: applied knowledge that solves problems. If a problem can’t be solved through technology, it’s not a technology problem: it’s a physics problem.

So in the future, cops will do absolutely nothing other than attempt to stop crime, and train to do that in the best possible way – unless social barriers prevent it.

Unless social barriers prevent it, predictive technology will show where crime is likely to occur, with very high accuracy. Some people might think that there’s no social barrier that should prevent such an obviously worthy goal. Some people will be more concerned about social harms that might come from errors and bias. Some people will be equally concerned, if not more, about the surveillance required to enable predictions. And yet some others believe that citizen surveillance could be a safer alternative to state-operated surveillance – or maybe that some combination of the two, formally or informally, would work optimally. But in any case, if it becomes known that a crime may be stopped, regardless of how it might be known, the police should be sent to stop that crime. Few people could possibly disagree that this would not be what we want from a perfect police force, which don’t forget, is perfectly trained.

As for the people who do all of the other things that police do today – some might argue that these are still police officers, that they are still as essential and honorable, if not even more so. And indeed, it is irrelevant whether or not they are called by the word “police” and irrelevant whether they wear a uniform and irrelevant where their paycheck comes from, from a technology point of view. Social factors determine whether they are called “police” or social workers, whether they are public or private or nonprofit. Those kinds of things have nothing to do with technology – although technology could certainly help determine which social choice is most likely to be optimal.

Social factors also determine whether those other “police” (whether or not so named) are allowed to carry weapons of any kind. None of these people are performing any tasks that are particularly likely to discover a crime in progress, so they clearly don’t need a weapon most of the time. Crucial exception: tasks that routinely involve interactions with victims, actual or potential, will of course discover crimes in progress. But as this is discovered from a victim, no weapon is needed unless for some reason the perpetrator is nearby, as is usually the case with domestic violence. Even in this case, it is clear that the task of ensuring safety is different from the task of preventing ongoing violence, so these are obviously separate jobs, only one of which is likely to need a weapon.

Social factors determine whether or not people who spend so much time doing social work should be able to carry any particular kind of weapon. Whether a “police officer” actually needs to carry a weapon is a social question. For example: maybe a political reason requires all the people doing all this driving around, talking to people, and filling out paperwork to be called “police.” And maybe other social factors require all people that are called “police” to carry weapons that they don’t need, for example for recruiting purposes (assuming that some people join the police at least in part due to their affinity for weapons). As a counter-example: maybe for political reasons, only the people who are actually trained to stop crime will be called police, and all of the other people will be some category of social worker (whether public or private). In that case, it seems unlikely that anyone would want the social workers to carry weapons. But it’s very clear from a technology perspective that only some types of work that we call police work today requires any kind of weapon.

So, in the future, the police as we know it will not exist, as a matter of technological fact – though this is endlessly contingent on social factors. In a perfect world, most people that we call “police” today would be doing the exact same thing that they do today, in terms of time, but they wouldn’t carry guns. Any rational person wouldn’t even want them, at least not for work, as they would know that they are unlikely to ever need to use them. (This is completely independent of any 2nd Amendment argument for or against carrying guns, as those arguments apply to all citizens, not just particularly to police.)

Like all technological predictions, the inevitable end of police as we know it is highly contingent on the expected operation of an extraordinarily complex and interrelated system of infrastructure and endpoints – but this is dependence on social infrastructure and people, not technology. Nevertheless, any good technologist should understand all relevant contingencies.

It’s very easy to imagine an attempt to reach this perfect world that inadvertently turns into a totalitarian police state enabled by technology – we’ve all seen those movies and shows many times now. It’s very tempting to imagine that enough social problems can be addressed so that technology has the social basis it needs to be successful – but there isn’t really much data that should give anyone optimism. So good technologists should spend most of their time finding data and implementing solutions that address the social infrastructure that is required for success.

I didn’t intend to include any moral suasion in this very dry essay, but I can’t help but end with it. Technologists: stop building weapons (anything that enables the police state), and do the social work (data and tools to solve the social problems that prevent us from working on more useful technology).

ETA: Someone suggested the perfect slogan for techies who want to reboot the police: CTRL-ALT-POLICE.