19
$\begingroup$

A common sight in science fiction is a police robot force. But as most people know, robots (even small basic ones) are expensive, which raises the question: why would they bother? Let's assume that the American government does have the technology and means of producing these robotic soldiers independently. The only factor is price and public appeal. (Naturally people would be intimidated) Never mind the flaws such as: Hacking potential, EMP devices and lack of morality.

Obviously there are bugs in the system, as is to be expected. What reasons would a country have for spending billions on having a robotic police force to replace a human one?

$\endgroup$
18
  • 5
    $\begingroup$ Worth noting: the "cost of raising a child" in the US is around $250k, and does not include college, just the bare minimum of keeping the child alive for 18 years. $\endgroup$
    – Cort Ammon
    May 26, 2016 at 5:39
  • 3
    $\begingroup$ @CortAmmon but a parent is going to be paying that whether or not the kid joins the force, while a police robots is a yes or no kind of deal. $\endgroup$
    – TrEs-2b
    May 26, 2016 at 5:46
  • 4
    $\begingroup$ Also: the police are not there to protect the people, but to be a tool for the government to influence people. With "ordinary people police" government can deduct their actions on the "human factor" of employees and with "robots police" - errors in the software. And so do whatever they want. $\endgroup$
    – Exerion
    May 26, 2016 at 5:47
  • 2
    $\begingroup$ Robots are expensive today, but in the shiny future they will be cheap. $\endgroup$ May 26, 2016 at 10:28
  • 3
    $\begingroup$ If it doesnt have to 100% robot, imagine a human cop with 3 robot-cop as aditional help. The human can instruct them, combining the human emotions and evaluation, and the power/shielding/weapons a robot can harbour. $\endgroup$
    – Martijn
    May 26, 2016 at 10:39

10 Answers 10

44
$\begingroup$

Why do robots ever get deployed for anything?

  1. When they are cheaper.
  2. When they are better (more accurate, etc.).
  3. When the work is dangerous.

Dangerous

So how does that apply to law enforcement? Let's start with the last point. Police work can be dangerous. It involves preventing people from doing what they want. Some people will react violently to this.

Better

A corollary of the danger issue is that robots can react less violently. One of the main reasons that police use violence is to prevent violence done to them. A robot could allow itself to be destroyed in a situation where a police officer might feel compelled to shoot. This makes the robots better in those situations.

Robots also have the potential to do things like on-the-spot analysis of fingerprints or forensic evidence that an officer might need a specialist to do. Not for the equipment but for the training. And of course robots are not emotionally impacted the same way that humans are. This allows them to respond more dispassionately. Robots can be controlled remotely more effectively than humans. Their communication protocols can be more efficient than verbal reports and orders. Live sound and video from the robot; efficient digital commands in response.

Robots follow their programming more than humans follow protocols. For example, the Freddie Gray case was caused in part by a failure to follow the official rules in securing a prisoner. A robot wouldn't have done that. Once their programming is updated, they will follow the new programming. They won't continue to do things the way that they always have. Robots also won't disable their cameras.

Cheaper

Humans cost a certain amount per year to hire. Don't forget the costs of pensions and insurance in that number. It's quite possible for a police officer to cost \$100,000 (US) a year even if only making \$20 an hour. If you can lease and maintain a robot for less, that allows the agency to either save money or increase their presence. A ten year, $3 million lease for a robot that operates 24x7 may actually be cheaper than hiring three or four officers for the same period.

Robots might integrate better with vehicles. Currently a motorcycle officer rides a motorcycle. The robot version might be an attachment. Robots might use such smaller vehicles rather than big, blocky, fuel-guzzling cars. Or an entire group might ride a single pickup truck.

Robots can power down when not in use. It's easier to maintain reserves that way. Activate them in response to calls. Humans would have to be given make work in a similar situation. You can't really tell them that you don't feel like paying them this month, not enough crime. But stick around, we may have more crime next month.

Other

Note that we already replace humans with robots in some circumstances. There are robots that disarm bombs now. They are currently rather stupid, being essentially remote operated. Traffic cameras can be considered robotic replacements of traffic officers. A parking monitor robot might make sense.

I thought that the show Almost Human did a good job of integrating androids into the police force. They took the risks and allowed single human officers to go about safely with android partners.

$\endgroup$
9
  • 3
    $\begingroup$ I really like your point about robots being able to use specialist tools on the fly, like DNA analysis. Also the point about transport reminds me of the first star wars film with the small droids all over the place, and packed into an APC. Presumably one could also airdrop in hundreds of android police by helicopters or airplanes quickly and cheaply to rapidly respond to situations like riots. $\endgroup$
    – user20787
    May 26, 2016 at 15:05
  • 3
    $\begingroup$ Don't forget that in a police state situation a robot's loyalty is unquestionable, whereas a human police officer might defect, or refuse orders. $\endgroup$
    – AndreiROM
    May 26, 2016 at 15:12
  • 1
    $\begingroup$ @inappropriateCode I'm not so sure. If we had such portable and automated equipment, why couldn't an officer do the same? It's totaly portable and automated, remember! Just point and shoot (or whatever). It's like the the scene in Short Circuit where the robot blows up a tank with a laser. Everyone else goes "Wow, robots are so powerful!" and I went "That power system and laser setup are amazing, they should put them on tanks." Robots are awesome, it's just that they need to be justified because they can use the accessories better, cheaper etc. then existing options with the same gear. $\endgroup$
    – wedstrom
    May 26, 2016 at 19:03
  • 3
    $\begingroup$ You talked about preventing physical damage to humans, but you may also want to mention psychological damage. PTSD could be severely reduced if we replaced humans with robots in military and emergency situations. $\endgroup$
    – Ethan
    May 26, 2016 at 19:06
  • 1
    $\begingroup$ You can't do DNA analysis on the spot. A single DNA sample takes months, and that's in lab settings. Fingerprint scanning is fairly valid, but I doubt it would lead to much reduction in the time taken to solve a case. Everything else is on the nose, however. $\endgroup$ May 26, 2016 at 21:33
18
$\begingroup$

"Naturally people would would be intimidated"

This depends on how popular the police are with the public at the moment. It's perfectly possible to make a racially discriminatory robot police force, but it's also possible to make one that isn't.

Robot police won't (unless specifically programmed to by evil overlords):

  • solicit bribes
  • discriminate on the basis of race or religion
  • get angry and assault suspects or deliberately apply excessive or lethal force
  • rape sex workers
  • steal cash or drug evidence
  • deliberately lose or destroy exculpatory evidence
  • lie about their actions
  • engage in coverups of illegal behaviour by the well-connected
  • fabricate crimes or search pretexts in order to meet quotas

Some places are already introducing body cameras in response to mistrust of the police. I can see robot incorruptible police being extremely popular if the existing human police are bad enough.

One aspect of this is that, even if not entirely bulletproof, shooting a robo-cop is property damage rather than attempted murder. Therefore there's less need for pre-emptive violence against suspects using the excuse of officer safety, and robo-cops would cause far fewer shootings.

$\endgroup$
4
  • 6
    $\begingroup$ Robot police won't [...] fabricate crimes or search pretexts in order to meet quotas Well, generally the problem with quotas and policing isn't that police fabricate crimes to meet the quotas, it's that the laws are bad, and would result in "everyone" being arrested or ticketed if they were actively enforced. That's one of the serious issues when law and AI or automation meet - real world laws are actually designed to be selectively applied by people, based on their judgement and many subjective factors, which doesn't translate well for computers/robotics. $\endgroup$ May 26, 2016 at 16:20
  • 2
    $\begingroup$ Related point: Robot cops are less likely to turn on the government if they do something that is unconscionable. If a government could field such and was planning something controversial, they'd be deployed. $\endgroup$
    – The Nate
    May 26, 2016 at 19:09
  • 1
    $\begingroup$ Unless hacked. Hacking a robot police officer would be hours of fun. $\endgroup$
    – Pharap
    May 27, 2016 at 6:39
  • $\begingroup$ I would not want these robots to exist if their entire software wasn't open-source. For me, it would be far too dangerous to live surrounded by robots, which are badly coded and waiting for someone to hack them, or with holes built-in by corrupt employees. $\endgroup$
    – Sebi
    Jul 18, 2016 at 10:41
4
$\begingroup$

But as most people know, robots [...] are expensive

If they are ubiquitous, one could assume the cost of robot-cops (totally not a trademark) is a solved problem. If you want to make a fair comparison, you must account for the price of training a human, their equipment, their retirement, their hospital bills or funeral, and maybe the checks sent to their families.

Why robots over humans?

You can train a robot much faster. If the robot dies, you can transfer its databanks to a new one. They are typically superior to humans in terms of strength, speed and reflexes. Intellectually, they can process more information, and access any data in the system on the spot, which is an interesting trait for a detective. They can also operate 24/7 without having to pay overtime.

Lack of morality is an interesting point, one that can also be found in humans. On one hand, there's the cold logic of a robot's programming, which drives them to make the "right" choice, depending on their programming. On the other hand, human can blatantly disregard human life. In a number of countries, police corruption is a big problem. You can't corrupt a robot.

Related to the previous point, robots have no second thought. If there is a danger, they are more likely to dive into it if it means saving a life, whereas a human would rightfully so fear for their own lives. In case of an injury, a robot can get repaired and get back in active duty much faster as well.

And because they have no strong feeling one way or the other, they might not kill a kid on sight because he's black.

Why not robots over human?

In the present, humans aren't really ready to trust robots with their lives. Likely, that doesn't change much in the future, so first contact would probably be cold. Human-cyborg relations would probably get better over time though.

Security would be a major concern, the possibility of your police force switching off or turning against you is a frightening one. However, as the hacking capabilities of bad guys increase, so does cybersecurity. You can take steps to mitigate the risks of hostile takeover. Ultimately, the common criminals will likely lack the proper skill to hotwire high-tech machinery so the real threat is organized crime, which is already known to bribe/threaten police officers. There is still no such thing as zero risk, so it comes down to how acceptable the risk is versus the risk of humans losing control of themselves.

Another thing to consider is that if they'll be chasing human criminals, then a human mind might be a better frame of reference. Humans are most likely to better understand how a human criminal think than a robot would.

Conclusion

A favorable environment I think would be high crime rates and/or high probability of getting your skin shot. In that case, a replaceable police force would make sense.

Another possible scenario would be lack of trust in human police. If your police force is known for its corruption, then a robot might be a better sight than a cop.

What I would recommend is to have humans and robots doing the job together, as opposed to one or the other. They both have strengths and weaknesses, so going one way may solve some problems while creating other.

A last point I'd like to make is that police is supposed to protect and serve the public. In an ideal world, financial considerations shouldn't matter as long as the public service is effective. But you know, that's ideally speaking.

$\endgroup$
4
  • $\begingroup$ "You can't corrupt a robot" - while this is far from true, the idea can certainly be sold that way. In reality, robots are susceptible to bugs, hacking, and manipulation by the people who make them or who have administrative access to their systems. $\endgroup$
    – komodosp
    May 26, 2016 at 10:21
  • 1
    $\begingroup$ @colmde It meant corruption in the exact same sense you would corrupt a police officer. Technical difficulties and hacking are addressed later. As for direct human intervention, it would also classify as hacking. $\endgroup$ May 26, 2016 at 11:19
  • $\begingroup$ I would have accepted this answer if you had not implied that all current police are prejudiced against one specific demographic. Oddly enough by adding the term "might not", you also imply that the robot "might" kill them just because they are that same demographic. $\endgroup$ May 26, 2016 at 13:41
  • $\begingroup$ @Keeta I'm only implying it's a thing that happened, which is true. Police brutality/discrimination is a known problem in a number of countries, not just the US. $\endgroup$ May 26, 2016 at 13:57
4
$\begingroup$

Many excellent points have already been made, I'd just add that although android police might be expensive, the alternative might be worse. Consider situations where the police force is corrupt. Yes, the USA is a good example given how American officers are too aggressive and racist, and will abuse their power with lethal effect a disproportionate amount of times. But that doesn't compare much to examples of the near total collapse of public institutions.

Take examples like Afghanistan, Mexico, Iraq, or indeed Northern Ireland during the Troubles. In many of these instances when the police can't be trusted the military is sent in, and they may actually make the situation worse.

After the American led invasion of Afghanistan a secular police force was established, but they quickly became extremely corrupt and were frequently accused of extortion, rape, and murder. Add this to the secular court system being slow and corrupt, and the result was that people began to support the Taliban insurgency because they believed they would offer more principled alternatives. Somewhat similarly when the Troubles started in Northern Ireland, the region's police force at the time, The RUC, was infamously sectarian and almost exclusively staffed by officers from the protestant community. Irish Catholics had no confidence in it, and so the British government sent in the army to restore law and order. It worked for a short time because Irish Catholics viewed them as less corrupt. But then Bloody Sunday happened, where Paratroopers shot dead unarmed protesters, and recruitment for the PIRA spiked and the violence intensified.

Mexico has a similar problem for different reasons. Mexican public sector workers of any worth, from police and soldiers to politicians and judges, are subject to a "silver or lead" policy by the cartels. Basically they bribe them or kill them. In this situation even when the government sent in the military to help the situation they couldn't be sure they weren't corrupt too. In fact one cartel known for its brutality, Los Zetas, was created when Mexican special forces trained with American counter-terrorism support defected to the cartels.

So not only would android police be less corrupt, aggressive, or outright traitorous, but they also don't have families who can be threatened to keep them under someone else's influence. And because they are all equally disciplined and impartial you can't just kill off the few good officers and then let a law of averages make the rest of the force susceptible to corruption. I imagine if you offered the people of Mexico, Afghanistan, Iraq, etc, an android police force tomorrow they'd jump at the chance. As with many black communities in the USA.

The main downside is the risk of software error, or of particular context being misunderstood by the androids, or them simply having no programming to handle an unusual situation, or these androids being hacked. A few years ago the Iranians managed to steal an American drone by hacking it, and brought it back to Iran; it was an excellent PR stunt, and exposed serious flaws in American military cyber security. The same sort of things would happen with android soldiers or police. But if the criminal are impoverished or don't have access to the training/technology required to hack it wouldn't be a major threat.

$\endgroup$
2
$\begingroup$

It would only be feasible to begin deploying a robotic police force in the event that the dangers to human police became so great that attrition - from casualties or resignation - of human police exceeded the rate of recruitment of new police when the recruits and experienced police are being paid an amount equivalent to the cost of purchasing and maintaining a police robot in the field, or when - despite an unchanged or even decreasing level of risk for human police officers - the cost of purchasing and deploying a police robot decreases such that it is cheaper to field robots rather than humans.

That is to say, when bots become cheaper than cops.

That said, at the probable cost of a reasonably effective robotic police officer, at present or in the near future, for a robot to be more cost-effective than hiring humans would mean that law and order must have broken down almost irretrievably. There would be many places where the police fear to tread in any strength. Murders of police officers would be a daily occurrence. Police on the street most likely would never operate alone or even in pairs, but might travel in groups of three or more. Police vehicles would be armoured against small arms and even RPGs. Everyone would own a gun or else live in a gated community guarded by private military units armed with military-grade weapons so that they wouldn't have to sully their hands by holding a gun. Emigration would be a major issue, with people leaving the nation for safer pastures.

Even then, deploying a robotic police force would be an almost stop-gap measure, as with that level of lawlessness, the nation's days are numbered, and it is likely that the government would become insolvent in relatively short order. It is likely that the nation deploying said robotic police wouldn't have the wherewithal to actually build them, but would be buying them from a more advanced, less lawless other nation.

However, the government of a nation in that much trouble is unlikely to spend that money on police robots, they're more likely to divert that money to line their own pockets. Only a truly idealistic government inheriting the aftermath of years of government corruption might actually spend that much money in that way - if they had it. Perhaps they executed the previous, corrupt regime and nationalised all their assets...

In the further future, as the price of robots decreases and their effectiveness increases, there may come a point where even a relatively peaceful human society would find it both cost-effective and politically neutral to replace human police with robots.

$\endgroup$
2
$\begingroup$

On robots being expensive - Technology is only expensive because of the research and development costs, which are once-off costs per technology. So "Robots are expensive" because they only produce them for very specific applications, and so only few of them are produced.

If you're mass producing, the price comes down significantly, as now you're just dealing with manufacturing / distribution costs + profit.

So the reasons to employ a robotic police force would be:

  • Widespread corruption amongst rank-and-file cops as well as mistrust of them.
  • Commonplace violence against them.
  • Common use of robots in everyday life - this will help both with the cost and acceptance by the public.
$\endgroup$
1
$\begingroup$

I can see a simple set of steps which might lead to a robotic police force.

1: First the military has a strong incentive to replace troops with drones and robots. A drone might cost a couple of million but photos of troops coming home in bodybags costs far more.

We're already seeing the transition in the american military. Drones are cheap and expendable compared with soldiers.

So it's only a matter of time before the military develops highly capable ground infantry capable of operating in urban environments.

So now you have a production line producing robots soldiers.

2: So the up-front costs of production are shouldered by the military which massively reduces the price for other consumers.

The companies making them want to expand their market share so they're likely to give some a new paint job and some software updates to try to sell them to the police. Is there a shootout where you're afraid that officers are going to die? Send in the robots! They're fast, they're politically expendable and they follow the law with a rigidity beyond that of any human. They're not going to put the boot in when the suspect is on the ground and they can follow procedure perfectly. The recordings from their cameras are even admissible in court when it comes time to try the gunmen.

Now we have some police forces with a small reserve of robots.

3: During staff shortages in some districts they're trialed for more routine use. Public complaints of abuse go down, conviction rates go up and even when mixed with human officers they're like badge-cams on steroids since all the other officers know they're being recorded.

4: Roll on a few years and more and more forces have realized that a highly robotic force has a lot of advantages. They still need some humans for the person-facing and management side of police work but more and more beat-cops are being replaced by robots. The unions hate it particularly after a string of convictions of officers for criminal behavior recorded by robot officers and in some regions there's massive strikes and conflict to try to resist their introduction. But they're popular with the public and ultimately most police forces end up going almost totally robotic.

$\endgroup$
1
$\begingroup$

Because the Government or ruling elite had reduced its citizens to binary digits and had no emotional connection to them. Therefore the Police would need to only react to yes and no conditions.

For example Q from Robotcop to Citizen. "Why are you talking to the zero binary digit. You are only allowed to converse with the binary digits in your own matrix. You have not been allocated additional memory to cope with the imagination necessary to understand what a zero binary digit is speaking as he comes from a world that is not controlled"

As our world becomes more controlled and at the same time the existing World Order is no longer able to direct behavior then it requires more force to prevent it collapsing.

Having a Police Force that could turn on its own Government as it tried to repress it citizens because they would not accept for example, a lower wage structure or forced movement due to climate change would not suit them.

Much better to have a robot-cop who will not listen to any discussion and will respond immediately with an arrest or the use of force should you be outside the permissible programmed parameters.

You only have to see how David Cameron interacts with his citizens to see that governments is slowly moving away from its people. The bigger the distance the less democracy we have and the more force they use to control us.

Even our still human Police have become more robotic in their interface with us. As soon as you disagree with them - then out come the cuffs. So how much better a robot cop who you cannot argue with.

$\endgroup$
1
$\begingroup$

While robots can be expensive people are really, really expensive.

The average cost to raise a child in the USA is about 1/3 of a million dollars. A pop. To maintain population, you have to supply your workers with enough money to raise their own kids at replacement rates (aka, a living wage).

If we throw on top of it the costs to maintain morale and food and shelter and post-work care for the humans, plus training and insurance and the like, the cost to employ a single human over a 30 year span can easily reach 5 million dollars. These humans have about 75% downtime (they only work 1 hour in four!), and their performance degrades and costs increase if they are run for longer.

If you manage 50% uptime on your robots and a 20 year lifetime, if maintenance + construction costs are as high as 7 million dollars they are going to be competitive even if they are no better or worse at the job.

The political cost of dead police officers is large, let alone financial, and training police officers to risk their lives to save the lives of citizens they interact with is difficult (even if you value both the non-police and police lives equally). Doing so with robot police may be easier, where you can risk a robot police officer "death" in order to ensure the safety of the public. This is similar to the advantage drones offer in the military: they can risk being shot down to fullfill their mission.

Politically, creating a manufacturing base beholden to your good graces is a great way to get contributions and reinforce both your parties and personal political and non-political income streams. Unions tend to be less effective at this, and union contributions as a whole tend to be party-aligned not just kickback-based: the members of the union often object to their money going to a party they don't personally support, while shareholders are held at a longer reserve. And very few unions make a practice of mass-hiring former politicians and senior public administrators after they retire or when they go on sabbatical.

So even if they are inefficient, there is a reason for politicians, think-tanks and manufacturers to "sell" the idea to the public.

Finally, if robots are capable of being a police officer, we are going to be close enough to strong AI that making most humans obsolete to those that command the heights of the economy is in sight. Feeding those millions of people who aren't providing services to those that own the capital is wasteful of good land that can be used as a hunting preserve, or a nature retreat, or just a good view out their window. An armed police force may identify more with the peons than the ownership caste, and a small number of technicians and industrialists may be suborned with wealth easier than a mass police force.

Even without that situation, a disarmed populance and an automated and armed police force could reduce the risk of coup or revolution.

So there are strong incentives to sell the idea above and beyond any actual practical advantages. The practical advantages, together with incentives to sell it, should provide a strong enough reason to deploy it when it becomes technically possible. A capital or state subsidized mass media can be directed to make the baseline assumption "of course this is better", and automation successes in earlier less "fraught" human-interaction tasks will normalize it for the populance.

$\endgroup$
-1
$\begingroup$

I would rather have a robot arrest me than a human. No human has the right to touch me or force me to do anything. A robot isn't human therefore they can and I will respect it.

$\endgroup$
1
  • $\begingroup$ Welcome to the site. Consider editing your answer to expand on these ideas; while this is something you can work with, it may help the OP if you broaden the idea and discuss it a bit more. $\endgroup$
    – Zxyrra
    Dec 24, 2016 at 4:46

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .