Robot's role in killing Dallas shooter is a first | Hacker News Reader

Robot's role in killing Dallas shooter is a first

money.cnn.com

67 pointsacjohnson5599 comments

It's worthwhile for students of American history to compare this to the apprehension of Lee Harvey Oswald. After Oswald shot and killed President Kennedy, he shot and killed a police officer. And when police cornered him at the Texas Theater, he fired at another police officer, but his gun jammed. Oswald was not executed in a hail of bullets. Instead, he was lawfully taken into custody.

John Hinkley, Jr., who shot President Reagan, also was not executed by police or the Secret Service. Indeed, he received due process and it turned out he was mentally ill, and he was incarcerated and treated accordingly.

Both McKinley's and Garfield's assassins received trials, and were executed after due process of law. Lincoln's assassin, John Wilkes Booth, was executed in a farmhouse by a soldier, Boston Corbett. Although Corbett claimed Booth drew a gun on him, Corbett's superior reported he shot Booth without pretext. Corbett was arrested until public opinion forced his release.

Even going back to one of the first attempted assainations of a President, a shooter who fired two pistols at Andrew Jackson both of which misfired, due process was given. The would-be killer was found not guilty by reason of insanity and confined to a mental asylum.

How far off are robots that can apprehend and restrain a gunman while minimizing the risk to the lives of both police and suspect?

It doesn't seem impossible, except for lack of a commercial motivation...

Pretty damn far, look at Atlas that's about as good as you get....

There is really no lack of commercial motivation the US army would be your 1st customer...

Neither Oswald nor Hinkley were holed up in a defensible position known to be sufficiently armed and inclined to shoot anyone who got anywhere near.

Oswald's gun jammed because the first approaching officer was close enough to try to grab it when Oswald pulled it out, and part of the officer's hand physically blocked the hammer. If Oswald had taken out the gun as soon as officers entered the theater and started shooting, I bet he would have been shot.

Also, the Dallas shooter claimed to have bombs in the area. That greatly increased the uncertainty in how much of a risk he posed. The Oswald and Hinkley situations had a much smaller uncertainty.

And what about public executions of plenty of unarmed colored civilians and even their kids? Justify it with "uncertainty of how much risk" everything on this planet poses?

> Oswald's gun jammed because the first approaching officer was close enough to try to grab it when Oswald pulled it out, and part of the officer's hand physically blocked the hammer.

Despite that level of uncertainty he wasn't shot by someone in uniform, arrested and was given due process.

That's unjust because they aren't holed up in a defensive position claiming that there are several bombs scattered throughout downtown.

Why would they kill him if they took seriously his claims that he'd hidden bombs? They would want to know where he had hidden them.

Fears of a remote detonator are not a real excuse either, he could could just as easily have a dead-man switch that detonates the bombs when he releases a trigger.

> Also, the Dallas shooter claimed to have bombs in the area

Sounds like an excellent reason to make every effort to take him into custody

As I understand it he was well-armed and well-armored, such that getting into a position to take him down non-lethally wasn't an option.

The robot/bomb solution does seem crude. How far off are we from robots who can get close enough to deploy some non-lethal method of disabling someone similarly armed.

You mean like the Terminator?

Robocop.

If they had a robot capable of taking in a bomb, could they have not used some other gas grenade option? To me, a bomb seems too likely to cause collateral damage (barring all other ethical concerns).

What sort of "gas grenade option"? Tear gas is only marginally effective. In reality there is no such thing as safe and effective knock out gas like you see in the movies.

https://en.wikipedia.org/wiki/Moscow_hostage_crisis_chemical...

"Safe", no. Better than a bomb, absolutely.

Moscow's was a special forces (think Delta/SEAL Team 6 level) operation. City police departments, even in Russia, do not have equipment nor training to wage chemical warfare on short notice.

Flash bang, smoke, Capsicum?

If he didn't have gas mask, a teargas grenade would cause a lot of discomfort and probably end a standoff.

Why didn't they use the robot to eg administer Thorazine?

Presumably the threat of using a trigger: exploding him and any electronic trigger he had probably amounted to the lowest risk according to their analysis. e.g. relatively low-tech assailant, some conviction they've actually planted IEDs. Perhaps they'd bluffed using a phone as a trigger.

I doubt he had hard armor plates that would stop a rifle bullet, and they're heavy, so their coverage is pretty small.

More likely there just wasn't a good shot to take of him, the only question is why not wait him out, and the threats of bombs and the fog of war (and this was much closer to war than any incident in recent memory, it's reminding me of the cultural '60s and those who wouldn't let it go into the cultural '70s) made it reasonable to take a more serious although not necessarily deadly methods (it's not like they used a fragmentation grenade) to stop the threat he potentially still posed.

And given his was a lot better at close quarters combat than his opponents in general (http://bearingarms.com/bob-o/2016/07/08/video-suggests-cop-k...), storming is position is not something I'd ask anyone to do.

Russia used an opium gas to disable terrorists in the 2002 theater raid.

It was a fentanyl analogue, and it killed several people, including hostages IIRC.

Is absorb fire and wait until their gun jams really the best option?

Absolutely not, and he had a SKS (https://en.wikipedia.org/wiki/SKS); military weapons, especially ComBlock ones, are built with tolerances to allow continued function as they get hot, generally at the cost of accuracy.

Lee Harvey Oswald is kind of crappy example, considering he was ultimately done in by Jack Ruby, a mobster.

Surely, Jack Ruby would never lie about his true motivation for killing Lee Harvey Oswald.

I mean, did Jack Ruby strike you as a dishonest man?

We're not talking about the expectations of mobsters, we're talking about the expectations of police, and Lee Harvey Oswald's interaction with them is relevant to that point.

Can you explain why you're comparing this to presidential assassinations and not other rampage killers?

Well the Lee Harvey Oswald case is in a similar situation of panic, right? After all the president got shot.

Though context matters, and what happened here seems to make sense, I've read a lot of cases where standoffs end in hailstorm of bullets, and others that end in prolonged negotiations. Yet both start off the same

Maybe the gunman did not want to be taken alive and behaved accordingly.

I think a better comparison would be the Orlando attack or the attacks in Paris which ended with the gunmen dead.

Wounded Knee.

I haven't yet been able to understand why using this explosive-carrying robot was "necessary" in this situation. It seems that they had the suspect cornered in a parking garage and, to me, that would seem like a good opportunity for some sort of non-lethal option. What am I missing here?

I'm sure someone will come up with a reason, but I doubt it will be more motivating than the expectation that police shouldn't issue death sentences - to me, at least.

If you object to the police using deadly force in this instance, do you also object to them trying to shoot the suspect?

If he is a credible and imminent threat (where imminent is immediate, requires similarly immediate action to protect oneself), then I wholly support neutralizing the threat.

I think that police are increasingly reaching for lethal force without exhausting nonlethal options.

I also think that, in this instance, he was a credible but nonimminent threat. So in that sense, yes I would object to shooting him.

A nonlethal response to continue containing the suspect, even at some risk to themselves, would have been more in line with the duty I believe police officers have. Minimizing the risk to themselves will not maximize the fulfillment of their duty (which is obviously true because the extreme is them never leaving the station, etc). I do not claim to know what the right balance is. However, I would argue that as public servants, their duty inherently necessitates taking on some level of risk in order to uphold our values of justice. Our values of justice require that they not kill suspects they are tired of dealing with - they must stand trial.

The fact that exercising non-lethal options against a currently active shooter exposes you to greater risk of being shot. You are of course free to argue that the police made the wrong call in the probabilistic tradeoff between their lives and the life of the shooter.

Given the circumstances of facing a trained active shooter who has killed and continues to try to kill officers, I would think that > 99% of cops would find the argument that you must preserve his life at the (high) risk to officers' lives incredibly insulting.

> they had the suspect cornered in a parking garage

Being cornered doesn't stop someone from shooting a gun. It especially doesn't stop someone who has a gas mask. It isn't unreasonable to suspect someone has a gas mask if he is firing at crowd-control officers in the sort of pop-pop-pause pattern reminiscent of the military.

Being cornered also doesn't stop someone from detonating a suicide vest.

EDIT: I would be interested to know how this comment doesn't contribute to the discussion, given the comment it is replying to.

But you don't get to kill someone just because they have barricaded themselves somewhere and threaten kill anyone who enters. They have to actually pose an ongoing uncontainable threat, i.e., a reasonable chance of escaping and going on to kill others. Otherwise, the police are obliged to wait things out, perhaps for several days, for the suspect to give up because of hunger/thirst.

I have no idea what the actual situation was, but it worries me that neither the reporters nor the police seem concerned with whether the suspect could have been contained. The rules governing the use of force are important.

Uncontainable is not part of their standards anywhere to my knowledge, and shutting down downtown for a week is not acceptable.

Deadly force is only allowed when it is the only reasonable way to prevent grave danger to others, and a contained lunatic is not such a threat. And the idea that police could or should execute people for obstructing downtown is bizarre.

EDIT: Even in the wake of the Columbine shooting, when police departments who had trained conservatively for hostage situations realized they needed to be more aggressive when dealing with active shooters (e.g., terrorists and suicidal killers), the Police Executive Research Forum emphasizes this isn't an excuse to go killing contained threats:

> Many [police department] policies note that Active Shooter protocols should not be used as a response to “barricaded gunman” situations. And some policies note that active shooter incidents are dynamic, and that an incident may go in and out of active shooter status in ways that could alter the police response. For example, a situation may begin as an active shooter incident, but if the shooter barricades himself in a room where he no longer has access to potential victims, and the police can secure that room and contain the shooter, the police response should shift accordingly.

http://www.policeforum.org/assets/docs/Critical_Issues_Serie...

I'd go further, it's only reasonably to prevent grave and _imminent_ danger to others.

The situation is kind of ironic, because it was vaguely associated with the protest against how police tends to kill people because they can't be bothered to find non-lethal ways to solve problems.

> Can't be bothered

Part of the problem in this broader conversation is a lack of knowledge by laypersons about what less-lethal options are available and their potential lethality. The most egregious example of this is when people (not on HN) suggest that cops should shoot suspects in the legs as a less-lethal option, as if the femoral artery doesn't exist.

I agree on the common misunderstanding about what less lethal options exist, but the justification is often misunderstood.

Rather, any potentially lethal force can only be used if there is a grave, immediate threat, and in this case the shooter ought to be compelled to take the most reliable action: shooting for the center of mass and attempting to fully disable the target. Shooting limbs is often not immediately disabling, and therefore wouldn't be an effective way of dealing with a life-or-death threat. Likewise, shooting for the head is rarely done by nonspecialists simply because it decreases the chance for success (not because the shooter wants to "give the target a chance to survive").

So you're right that if someone thought shooting limbs never resulted in death, they might mistakenly think it was an appropriate non-lethal option. But I think, in fact, most people are aware death is possible. It's just that they mistakenly reason that if limbs shots are less likely to kill then they ought to be a preferable method for disabling a target.

surely it's a well-reasoned tactical decision and not, "we're not letting this filthy cop-killer come out alive," because no one in our well-trained, civilized police forces thinks like that, right?

An argument could be made that seeing one's colleagues get shot by someone biased the cops' perceptions of "is this person still an ongoing threat?"

That's a risk you can marginalize fairly easily. You know you're sending in the robot in advance of the shooter, so you could easily consider preparing an explosive back-up option in case the nonlethal option fails.

>Given the circumstances of facing a trained active shooter who has killed and continues to try to kill officers, I would think that > 99% of cops would find the argument that you must preserve his life at the (high) risk to officers' lives incredibly insulting.

The officers' do not have the right to execute people just because they are too stupid, too careless, or too insulted to act with honor. You can't pretend to know a persons justification for their actions. The shooter may have had mental problems. The shooter may even have been coerced. You don't know the magnitude of their mistake because you didn't have a trial in which they could try to justify their actions.

And though it is unlikely, it is still presumptive to assume there is no reasonable justification for the shooter's actions.

> It especially doesn't stop someone who has a gas mask. It isn't unreasonable to suspect someone has a gas mask if he is firing at crowd-control officers in the sort of pop-pop-pause pattern reminiscent of the military.

It is definitely unreasonable. That's basically like a just-so story; a ad-hoc justification pretending a thing you want to be true is true just because you can imagine a way in which it could be. How someone is firing a firearm has very very little to do with whether they have access to a gas mask, and there are a trillion ways in which this kind of assumption can be wrong, and most of them are blatantly obvious.

> The officers' do not have the right to execute people...

I agree. If we establish that the officers perceived that they have control of the situation, then this paragraph becomes relevant. My current understanding is that they thought that they were facing a trained, prepared, still-active threat whom they were close (but not complete) to containing.

> it is still presumptive to assume there is no reasonable justification for the shooter's actions.

I don't think one could possibly build any sort of law-enforcement organization whose members believe there is reasonable justification for being pre-meditatively shot.

That's why it's not the job of law-enforcement to determine what reasons people have for breaking the law. They only must assume that they have some potentially valid, rational reason do so, and thus defer to a court system to determine what these reasons are and whether they are justifiable.

I wouldn't say that an active shooter should be kept alive at all costs, it just seems that if robots exist that can deliver an explosive device to the shooter, a robot could also deliver a tranquilizer dart to the shooter.

In addition to ethical implications, keeping your suspect alive also has practical advantages for information-gathering purposes. As it stands now, we'll have to settle for interviews with friends/family, social media autopsy, etc.

As I said in my original comment, I just don't understand how the situation got to the point of "we have no choice but to blow him up with our robot." I certainly realize that there are circumstances in which that would be the best option--news accounts that I've read so far just don't seem to justify it.

> I just don't understand how the situation got to the point of "we have no choice but to blow him up with our robot." I certainly realize that there are circumstances in which that would be the best option--news accounts that I've read so far just don't seem to justify it.

Agreed. The distinction isn't between robots and guns per se. It's that lethal force is only allowed when someone poses a grave and imminent threat, and it's hard to imagine such a situation where a slow robot is useful.

The slowness of the robot is a good point to bring up and a pretty strong counterargument because it indicates that the robot was only useful in a situation where the police perceived[1] that they had control of the situation.

[1] remember, if we're talking about what procedure police should follow, we have to think about the perceptions that they are responding to.

1) Tranquilizer darts aren't magic. Without calibrated dosage, the effects can range from nearly nothing to fatal. [1]

2) Tranquilizer darts work against unprotected skin, not heavy clothing and certainly not against body armor.

3) Where would one get a robot armed with a tranquilizer dart launcher? They're not exactly off the shelf items.

[1] https://www.quora.com/Why-dont-cops-use-tranquilizer-guns-wh...

The specifics of this situation matter less than the precedent it has set for future police actions around the world.

From now on, when faced with difficult or dangerous situations, police forces everywhere will at least consider using robots, because "Dallas police did it, and there was neither public outrage nor political fallout."

And following from that I guess robots doing traffic stops in high risk black neighborhoods.

The only positive I can think of, from this escalating into totally unimaginable madness is the most cities dont have the budget for all this sci-fi shit.

I can't even believe Dallas had a bomb placing robot lying around.

> I can't even believe Dallas had a bomb placing robot lying around.

That isn't uncommon actually. One common approach to handling potential unexploded ordinance is to evacuate the area and have a controlled detonation nearby. If you hear one boom, it was nothing. If you hear two booms, it was a bomb. Robots are often used for this.

Source: This is coming from my memory as part of MIT's Community Emergency Response Team back in 2012 & 2013. We didn't receive EOD training of course. We just did first aid training and crowd control at a few events, but we also go to sit in on discussions of EOD procedures and a tabletop excercise of an active shooter scenario.

Not to mention that there are police departments in the Dallas area which have drones.

It's important to remember that the shooter was specifically targeting police officers over what he viewed as massive injustices towards his own community. Using a robot with a weapon of mass destruction to execute the suspect does nothing to defuse that frustration and anger in others who may be sympathetic to what the shooter was trying to accomplish. In this case, two wrongs very much don't make a right. Unfortunately I'm not sure most citizens understand the historical significance of this event or the fact that the slippery slope is now well greased. How many more years till our police officers are armed with FPV multi-rotor drones capable of seeking out suspects and detonating explosives to take them out?

Not really a response, just venting: If you saw someone actually gun down a couple of your colleagues would you treat them as just a "suspect" or would you treat them as a murderer. I know why it's put like this but in some cases I find it a bit ridiculous.

This is the exact reason we have a justice system with courts and rule of law rather than supporting vigilante justice. If you remember, early on in this incident, a completely innocent individual was identified as the possible shooter with many absolutely convinced he was.

https://www.washingtonpost.com/news/morning-mix/wp/2016/07/0...

That's some George W. Bush style exaggeration. A small conventional explosive is in no way a WMD.

Yes, I intentionally called it a WMD because if you look at what law enforcement has been doing since 9/11, it's not uncommon to arrest someone in possession of a pipe bomb or flash-powder as an example on WMD charges.

http://abcnews.go.com/US/russian-citizen-pa-charged-wmd/stor...

> Bielat's ground robots, which are designed for law enforcement, aren't designed to be used for lethal purposes. But that doesn't stop them from being used that way.

I think this is exactly why companies like Apple are afraid of working with the gov - you just don't know how the gov will behave with what you give them once it's in their hands.

Whereas, when criminals use what you give them, few people will blame you. It's actually slightly strange.

It's strange to hold our government to a higher standard than criminals?

No, it's strange to hold companies selling to the government to a higher standard than companies selling to the public.

Doesn't seem that strange, with great power comes great responsibility

Knowingly selling to criminals on the other hand...

I've never been comfortable with the use of the term "robot" to describe a radio controlled car (or plane, boat, whatever).

Using a ruggedized R/C car to deliver an explosive is no different from a police officer with a good throwing arm tossing a grenade (or whatever) at the attacker. In other words, the action is completely controlled and executed by a person, not the machine.

Now, if this thing had autonomous "search and destroy" capabilities and the person's role was to identify the target area and approve the kill, the term "robot" would certainly apply.

I don't like the association being created between robotics and killing. In this case a person was 100% in charge of killing the suspect. The machine was not operating autonomously.

Just like the folks who attached a gun to a quadcopter. Where that thing went and when it pulled the trigger, save a malfunction, was 100% remotely controlled. Calling it a "drone" or "robot", which implies a level of machine-based decision making, would be wrong.

Agreed, which is why I always disliked the show "Robot Wars". It sounds better though than "Giant Armored R/C Vehicle Wars".

Indeed true. However, it boils down to the "video game UI makes killing easier" discussion. Which in this case, as far as I am concerned, is not the discussion to have. I think it actually distracts from the discussion we should have: Why do so many police interactions end with casualties in the US, compared to other countries. Obviously many think it's racism, and that may well be true, but I think it's also bad training and maybe a wrong overall approach to policing.

It was a remote controlled device, not a robot. It had no AI or built-in intelligence, and made no decisions. An officer at the controls is responsible for the shooter's killing.

Sadly in the minds of the public there is no distinction. Something which has always bugged me.

Perhaps. But the fact that they used a remote-controlled device is significant. It's quantum change in tactics for the police. It gives police far greater powers against suspects.

How long before they use drones to shoot suspects?

What about automated traffic stops, where a device asks you to roll down the window and scan your license and insurance?

If there's a robbery or burglary in progress, wouldn't it always make sense to send in a device instead of exposing an officer?

What about a bunch of kids on the corner who shouldn't be there? Why expose a cop to them, when one of them might have a gun? Instead, roll up in a well-protected car and speak to them through speakers and microphones and video.

I'll bet a remote-controlled device could eventually cuff them and take them into custody as well.

Don't underestimate the effect this is going to have on civil liberties.

It might also go the other way. Many police killings are due to jittery officers in dangerous situations. Take that out of the picture and you might get better policing and fewer suspect deaths.

So let's drop the semantic distinction and call it a drone. A drone deployed domestically against a US citizen.

Only remotely operated flying vehicles are drones in the eyes of the American public.

Why didnt they use some some gas agent, like russians did in the theater? (Without hostages this time, the risk of collaterals is removed).

Sending a bomb on wheels makes it look like just revenge.

That gas agent (allegedly aerosolized fentanyl [1]) didn't actually work so great, and wound up killing people anyway, so might've had the same outcome, but with more effort and just as much controversy.

[1] https://en.wikipedia.org/wiki/Moscow_hostage_crisis_chemical...

It failed because they used an anesthetic and it will certainly kill you if you can't control the dose. In this case they had 1 target who they were carefully monitoring. Once it reached its effectiveness they could cut it off.

That's not how anesthetics work. It's impossible to deliver a controlled dose or measure effectiveness just by spraying some gas in a general direction. Ask an anesthesiologist.

"Uh, mom?"

Seriously, she was a nurse anesthetist who retired decades ago when she had me, and she's mentioned how it's a somewhat precision effort. They monitor the patient to see how it's effecting them, since we're all different.

Yes they monitor the patient in an operating room using multiple instruments. And the anesthetic is delivered in a precisely calibrated dose based on the patient's size using a mask or IV. Which is completely different from the actual situation in Dallas.

Edit: sorry I didn't realize that the parent post was essentially agreeing.

(Uh, I wasn't the one to suggest that, seeing as how I knew from my mother these details. Sorry for my lack of clarity.)

That situation could have been handled better (from what I've read the deaths were all preventable by proper treatment), but even 16% lethality compares pretty favorably with the 100% lethality of a bomb.

A police force in the United States wouldn't stock anything remotely similar to that Russian gas.

On the other hand, nasty tear gas is available. Someone decided that trying to flush that guy from his position, such that he'd have to either come out with his hands up or come out shooting, was not desirable.

This wasn't a robot - it was a remote controlled vehicle. I do share the sentiments of others here that there should have been a non-lethal option to disable the perpetrator.

"The suspect stated that we will eventually find the IEDs." It is likely the talk about IEDs that brought the robot out in the first place. How the decision was made to use the robot to blow up the guy is something else entirely.

1. http://www.aljazeera.com/news/2016/07/dallas-police-suspect-...

Hopefully not a pioneering incident for the robotic enforcement of citizens.

The CNN video says "they're used to diffuse bombs..."

I don't think so. That would defeat the purpose.

"Defuse" maybe?

Not 3-Law safe

I dont think police are paid enough to risk their lives against someone who is actively trying to kill them. How much would you want to approach a situation like that in the way you are all detailing?

Militaristic police abuses cause kickback by oppressed citizens, so the police escalate with killer robots.

Seems legit.

Isaac Asimov's "Three Laws of Robotics"

1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.

2. A robot must obey orders given it by human beings except where such orders would conflict with the First Law.

3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

At the end of Foundation and Earth he added a "zero" option:

0. A robot may not injure humanity, or, by inaction, allow humanity to come to harm.

But he entire point of Asimov's Three Laws was to be a literary device for stories about the Three Laws being a bad idea. They weren't intended to be taken seriously, they were intended to illustrate the unintended consequences that can arise from the hubris of trying to codify morality in the first place. Robots governed by his laws in universe harmed people all the time.

If we ever do want to tackle the issue of the morality that governs AI, Asimov is the last person we should refer to as a credible source of inspiration.

I disagree, when we industrialize killing and mayhem with seemingly little consequence, nothing good will come out of it.

As for Asimov, I think he was on to something; your milage obviously varies.

> Asimov is the last person we should refer to as a credible source of inspiration.

Wait, why wouldn't we want to look at the failure modes he pointed out?

I don't think there's anything wrong with that, but I do think it's a mistake to take them seriously, except as an example of what not to do.

Right. Speculative fiction is just that: speculation. It is useful as speculation.

The three laws are a literary device and not an engineering proposal.

I think both Sci-fi and art has always been a "engineering proposal" from Leonardo di ser Piero da Vinci to Star Trek. The Smithsonian Institute seems to agree:

http://www.smithsonianmag.com/science-nature/ten-inventions-...

The Asimov stories that introduce the laws are exactly about their inadequacies and internal contradictions. They are cleverly chosen to seem plausible at first glance but cause endless problems when fleshed out.

Reply on news.ycombinator.com