Not fully automated. These may be autonomously driving, but not autonomously shooting. The article hints that that is technically possible, but also states that it is not implemented.
Can't we just get a kind of programmer's Hippocratic oath going here? Let's just all not facilitate autonomous shooting. I know both sides of the argument. I know it won't prevent it being done. I just mean if enough people agreed not to do it, then we'd at least generate enough disapproval for it not to be the subject of any PR Puff.
Seriously. Incremental improvements are worthwhile.
In general, militaries never want to comprise their effectiveness. So if given the chance to implement a life-saving intervention that harms their effectiveness, they will be reluctant at best to implement it. But this is a life-saving intervention that possibly even increases their effectiveness!
That's what truly makes it awesome. A way to save lives that militaries are not only willing to implement, but rather are actually enthusiastic about implementing.
That's a best case scenario. What it could do is make military intervention an even easier decision. The loss of human life is a deterrent to engaging in war. If one side isn't concerned about that, wouldn't they be more inclined to go to war?
Drone operators refer to children as “fun-size terrorists” and liken killing them to “cutting the grass before it grows too long,” said one of the operators, Michael Haas, a former senior airman in the Air Force
Conversely removing the threat to human life may reduce the probability of escalation and make a response more measured.
For example two men break through a security fence. A robotic patrol would seem to be much less likely to respond with deadly force than a human patrol would in that situation.
If the men were refugees or wounded soilders looking for medical care the more measured response of the robotic patrol may prevent a death and actually de-escalate conflict.
Also assume that the two men were hostile and fired on the patrol. The destruction of the mech is much less likely to provoke a political and civilian response than the death of a human and may lead to a more measured response.
My bigger concern is that a fully mechanized force may decrease the apparent cost of oppressing a civilian population.
Already, in the case of Israel vs. Palestinian resistance, we're talking teenagers throwing rocks and people firing WW2-class rocket artillery, versus one of the most high-tech and well-funded armies (with very good, uncontested air support) in the world.
With robots involved it would most definitely be a one-sided reduction in casualties, but it already is so one-sided I'm not sure that it would meaningfully alter the numbers in that conflict. Take the 2014 Israel-Gaza war, for instance (numbers from Wikipedia):
Israel: 70 casualties, 90% military.
Gaza: around 2 500 casualties, 30-40% military (depending on source, but even Israel estimates <50%).
Worth mentioning: the side that has the biggest causualities is the one who keeps attacking first. The only reason why more Israelis aren't hurt seems to be that they put a lot of effort into shielding their civilians.
There is little doubt that if stabbings, rocket attacks etc stopped Israel would stop as well.
... and, there is also little reason for doubt that if Israel stopped first it wouldn't stop attacks against them, quite on the contrary there is reason to believe they would escalate if left alone.
PS: while I feel this was worth mentioning I was not the one to downvote you, I think you argued reasonably.
>Worth mentioning: the side that has the biggest causualities is the one who keeps attacking first.
>There is little doubt that if stabbings, rocket attacks etc stopped Israel would stop as well.
Every violent action by both sides is framed as a retaliation for previous wrongs.
It's almost impossible to prove conclusively who started the tit for tat in almost all cases. However, if you do an analysis of most media stories or news reports in the US, there's a curious degree of unanimity about it.
Dismissing the proportionality of the responses (which you can measure) in favor of focusing of "who started it?" (which you can't), is effectively prioritizing propaganda over facts.
I assume that these devices will mainly being deployed against enemies not having those. Main advantage: no chance of mercy. Extremely valuable in civil wars...
Do note that the robots described in this article only drive autonomously, and when armed, need a human to (remotely) fire the weapons.
If the robot shot autonomously, you'd have a point about mercy. But the way it is described here, it's exactly the same mercy-wise as if the soldier manning the weapons was riding along in the vehicle.
Sounds good but the problem is, once all the robots of one side are killed, the winning side robots will attack and start to kill the losing side humans.
The issue isn't robots versus robots of course, it's robots versus humans. Robots are far less squishy and compassionate than their human friends.
Now if there were a treaty stating that all wars are to be fought in a robot battle arena, and the event was available to stream on Netflix, count me in.
We are fighting asymmetric wars where one side frequently doesn't have the resources or knowledge to use robots.
These robots will make it easier to kill people by reducing your own risk. I doubt they will reduce the amount of lives lost. At the very least the calculus is quite a bit more complicated. I doubt we will see robots fighting each other …
This comment does not make sense -- it isnt like the robots will be shooting at Clay pigeon -- the robots will be killing humans. Right now, most of the targets end up being civilians, that can only go up.
Weapons that autonomously decide when to explode have existed since WW2 with radar proximity fuses and modern guided weapons can select targets (e.g. "fire and forget" Brimstone missiles).
Those are very good points. I totally agree with you on mines.
But I would say that deciding to fire an autonomous missile in a hostile environment is not the same as deciding to unleash an autonomous fighting system.
In the first case, a human decides to unleash the autonomous missile once during a single event. Whereas in the second case, you are unleashing a system that decides for itself when to fire for every situation it is placed in. The second system is placed in many kill/no-kill situations and might not have enough discernment to realize that in this one special case the person in front of it is not hostile even though it is in a conflict zone.
An autonomous fighting system is likely to reduce casualties on both sides, especially in this case.
When you have manned patrols they engage the enemy because otherwise they would be killed or hurt, they have more incentive to prosecute any enemy combatants because they are a direct threat.
Autonomous systems do not have that, they can can engage with both lethal and less than lethal options but they do not have the same sense of urgency to prosecute the targets.
Additionally in this conflict specifically you have one side which wants to inflict casualties at all costs including it's own men and if possible take hostages even if it's just bodies because Israel has released 1000's of prisoners before just to get bodies.
If there are no manned patrols organizations like Hamas and Hezbollah would be less inclined to execute attacks on the border their attacks have no strategic or tactical value other than to inflict as many casualties on both sides as they can.
Any Israeli they capture or kill is a political win for Hamas internally, any Palestinian Israel kills is a political win for Hamas internationally.
Israel has been operating UGV's for 5-6 years now the Gaurdium went in service a few years ago https://en.wikipedia.org/wiki/Guardium the program has been "terminated" but it looks like what they've done is simply taken the drive kit from the Guardium and put it on the Ford 150/250/350 trucks they use because it's cheaper than making UGV's from the ground up.
If the patrol that Hezbollah attacked in 2006 was a "robotic" one, Israel might have retaliated against Hezbollah targets in Lebanon but it would not have invaded, the only reason it did was because Hezbollah captured the bodies of Israeli soldiers. Captives whether they are dead or alive is simply unacceptable to the Israeli military and public at levels people in the west simply cannot grasp.
Every body and every body part must be retrieved this is prosecuted with almost fanatic level of commitment (case and point the philadelphi corridor incident where Israeli soldiers spent 2 days under fire sifting through sand to retrieve body parts after an APC carrying munitions exploded), the IDF through its all of is operational history and all of its wars only has had 5 soldiers that were classified as MIA and their bodies were never retrieved.
This includes 3 tank crew members which were captured during the battle of Sultan Yacoub (Israel finally managed to get the tank back from Russia this year), a navigator that was captured by AMAL (a Lebanese paramilitary organization) and one soldier that went MIA during his conscripted service, this isn't combat related (suspected suicide/criminal activity).
That's it, every other casualties and captive since the war of 1948 has been retrieved, when you take that into account replacing patrols with Ford pickup's that Israel will not start a war over makes a lot of sense.
What's so impressive, technically, about a four wheel drive moving to a series of waypoints and shooting anything that moves and doesn't squawk friendly?
I'm surprised this hasnt been done sooner. Waypoint navigation is a consumer product, and we've had military grade target tracking for ages.
"Overkill" is the defining feature of Western military. We've got to the point where we consider a ratio of 1 (of ours) : 1000 (of theirs) unacceptable. It's very hard to present an argument that says we should accept more deaths on our own side in order to be "fairer" to the other side. However, the net effect is that we continue to minimize our risk in ways that de-humanize conflict and increase collateral damage.
I'm not sure we are really increasing collateral damage, relative to where we started 100 years ago. The days of firing 1-3 million heavy artillery shells in a day[1] are past, as are the days of dropping tens of thousands of tons of bombs indiscriminantly on a city[2]. Blowing up a single house with a drone-launched missile, or taking out a bridge with a single laser-guided bomb is far more surgical. One can imagine in the not-too-distant future bug-sized drones that could pick a target out of a crowd and inject a trace amount of some very deadly toxin or sedative.
It's true we are making less collateral damage then in the golden days, but the problem is that the more advanced one side becomes the less of a chance the other side has.
It's not just about fighting terrorists. We don't know who will be in command next. Giving this tech into the wrong hands will do some massive damages. There is always a chance to find ourselves on the wrong end of the drone.
Also the drone targets a single house, bombs it to oblivion and then someone says they are 80% sure, they bombed the right house with the right people in them. Not to mention a lot of times the people killed are civilians who were actually supportive, but because they are dead now, we just make the rest angry and make them side with the enemy.
We are a long way away making sure we are killing the right target.
There may be inconvenient and/or political repercussions - but those can be worked around by further magnifying the power imbalance. (Yes, I'm a cynic.)
30 Use the proof of real life usage in a low boil conflict to sell the technology at international arms fairs.
40 Use realistic test conditions to refine & develop new technology for 2 years.
50 GOTO 10
This efficient QA/development cycle is a serious competitive advantage, since it's hard for other countries to replicate the same controlled conflict conditions. Hence these great sales figures: https://rwer.wordpress.com/2016/01/12/the-15-largest-arms-ex...
A potential advantage or military robots is you could make them non or less fatal. In Aleppo type situations robots might be able to taser and arrest people instead of trying to kill them.
54 comments
[ 3.6 ms ] story [ 120 ms ] threadWar sucks, and I do wish that we could have world peace. But if war is going to happen, I'd rather it be robots that are the ones in harm's way.
This is already happening.
Seriously. Incremental improvements are worthwhile.
In general, militaries never want to comprise their effectiveness. So if given the chance to implement a life-saving intervention that harms their effectiveness, they will be reluctant at best to implement it. But this is a life-saving intervention that possibly even increases their effectiveness!
That's what truly makes it awesome. A way to save lives that militaries are not only willing to implement, but rather are actually enthusiastic about implementing.
Drone operators refer to children as “fun-size terrorists” and liken killing them to “cutting the grass before it grows too long,” said one of the operators, Michael Haas, a former senior airman in the Air Force
https://theintercept.com/2015/11/19/former-drone-operators-s...
For example two men break through a security fence. A robotic patrol would seem to be much less likely to respond with deadly force than a human patrol would in that situation.
If the men were refugees or wounded soilders looking for medical care the more measured response of the robotic patrol may prevent a death and actually de-escalate conflict.
Also assume that the two men were hostile and fired on the patrol. The destruction of the mech is much less likely to provoke a political and civilian response than the death of a human and may lead to a more measured response.
My bigger concern is that a fully mechanized force may decrease the apparent cost of oppressing a civilian population.
We always assume, that the technology always stays with the good guys.
This stuff is literally a dictators wet dream.
With robots involved it would most definitely be a one-sided reduction in casualties, but it already is so one-sided I'm not sure that it would meaningfully alter the numbers in that conflict. Take the 2014 Israel-Gaza war, for instance (numbers from Wikipedia):
Israel: 70 casualties, 90% military.
Gaza: around 2 500 casualties, 30-40% military (depending on source, but even Israel estimates <50%).
There is little doubt that if stabbings, rocket attacks etc stopped Israel would stop as well.
... and, there is also little reason for doubt that if Israel stopped first it wouldn't stop attacks against them, quite on the contrary there is reason to believe they would escalate if left alone.
PS: while I feel this was worth mentioning I was not the one to downvote you, I think you argued reasonably.
Every violent action by both sides is framed as a retaliation for previous wrongs.
It's almost impossible to prove conclusively who started the tit for tat in almost all cases. However, if you do an analysis of most media stories or news reports in the US, there's a curious degree of unanimity about it.
Dismissing the proportionality of the responses (which you can measure) in favor of focusing of "who started it?" (which you can't), is effectively prioritizing propaganda over facts.
If the robot shot autonomously, you'd have a point about mercy. But the way it is described here, it's exactly the same mercy-wise as if the soldier manning the weapons was riding along in the vehicle.
Now if there were a treaty stating that all wars are to be fought in a robot battle arena, and the event was available to stream on Netflix, count me in.
These robots will make it easier to kill people by reducing your own risk. I doubt they will reduce the amount of lives lost. At the very least the calculus is quite a bit more complicated. I doubt we will see robots fighting each other …
Where do we draw the dividing line?
There was still a human that chose the target, and a human that chose to fire on the target.
https://en.wikipedia.org/wiki/Brimstone_(missile)
This kind of ability to pick specific targets has been around for a while - e.g. SADARM
https://en.wikipedia.org/wiki/Sense_and_Destroy_ARMor
Edit: Of course someone decides to fire these things, but isn't that similar to someone deciding to unleash an autonomous fighting system?
Edit2: And what about mines?
But I would say that deciding to fire an autonomous missile in a hostile environment is not the same as deciding to unleash an autonomous fighting system.
In the first case, a human decides to unleash the autonomous missile once during a single event. Whereas in the second case, you are unleashing a system that decides for itself when to fire for every situation it is placed in. The second system is placed in many kill/no-kill situations and might not have enough discernment to realize that in this one special case the person in front of it is not hostile even though it is in a conflict zone.
When you have manned patrols they engage the enemy because otherwise they would be killed or hurt, they have more incentive to prosecute any enemy combatants because they are a direct threat.
Autonomous systems do not have that, they can can engage with both lethal and less than lethal options but they do not have the same sense of urgency to prosecute the targets.
Additionally in this conflict specifically you have one side which wants to inflict casualties at all costs including it's own men and if possible take hostages even if it's just bodies because Israel has released 1000's of prisoners before just to get bodies.
If there are no manned patrols organizations like Hamas and Hezbollah would be less inclined to execute attacks on the border their attacks have no strategic or tactical value other than to inflict as many casualties on both sides as they can.
Any Israeli they capture or kill is a political win for Hamas internally, any Palestinian Israel kills is a political win for Hamas internationally.
Israel has been operating UGV's for 5-6 years now the Gaurdium went in service a few years ago https://en.wikipedia.org/wiki/Guardium the program has been "terminated" but it looks like what they've done is simply taken the drive kit from the Guardium and put it on the Ford 150/250/350 trucks they use because it's cheaper than making UGV's from the ground up.
If the patrol that Hezbollah attacked in 2006 was a "robotic" one, Israel might have retaliated against Hezbollah targets in Lebanon but it would not have invaded, the only reason it did was because Hezbollah captured the bodies of Israeli soldiers. Captives whether they are dead or alive is simply unacceptable to the Israeli military and public at levels people in the west simply cannot grasp.
Every body and every body part must be retrieved this is prosecuted with almost fanatic level of commitment (case and point the philadelphi corridor incident where Israeli soldiers spent 2 days under fire sifting through sand to retrieve body parts after an APC carrying munitions exploded), the IDF through its all of is operational history and all of its wars only has had 5 soldiers that were classified as MIA and their bodies were never retrieved.
This includes 3 tank crew members which were captured during the battle of Sultan Yacoub (Israel finally managed to get the tank back from Russia this year), a navigator that was captured by AMAL (a Lebanese paramilitary organization) and one soldier that went MIA during his conscripted service, this isn't combat related (suspected suicide/criminal activity).
That's it, every other casualties and captive since the war of 1948 has been retrieved, when you take that into account replacing patrols with Ford pickup's that Israel will not start a war over makes a lot of sense.
You have people pretending to build a tank on one hand and people with automated robots on the other? Seems a bit overkill and pointless IMHO.
Of course it's quite impressive! Israel's R&D efforts are nothing short of that.
are also not made in an vacuum. substantial assistance and sharing with the US has certainly accelerated their abilities.
I'm surprised this hasnt been done sooner. Waypoint navigation is a consumer product, and we've had military grade target tracking for ages.
[1] http://www.longlongtrail.co.uk/battles/battles-of-the-wester...
[2] https://en.wikipedia.org/wiki/Bombing_of_Hamburg_in_World_Wa...
It's not just about fighting terrorists. We don't know who will be in command next. Giving this tech into the wrong hands will do some massive damages. There is always a chance to find ourselves on the wrong end of the drone.
Also the drone targets a single house, bombs it to oblivion and then someone says they are 80% sure, they bombed the right house with the right people in them. Not to mention a lot of times the people killed are civilians who were actually supportive, but because they are dead now, we just make the rest angry and make them side with the enemy.
We are a long way away making sure we are killing the right target.
There may be inconvenient and/or political repercussions - but those can be worked around by further magnifying the power imbalance. (Yes, I'm a cynic.)
https://www.youtube.com/watch?v=_mqDjcGgE5I
We make them ourselves, against ourselves.
Can't wait to see someone hacking these machines and turn them against their former owners.
20 Run live tests on technology
30 Use the proof of real life usage in a low boil conflict to sell the technology at international arms fairs.
40 Use realistic test conditions to refine & develop new technology for 2 years.
50 GOTO 10
This efficient QA/development cycle is a serious competitive advantage, since it's hard for other countries to replicate the same controlled conflict conditions. Hence these great sales figures: https://rwer.wordpress.com/2016/01/12/the-15-largest-arms-ex...
And it's been ~2 years since the last conflict, so...: http://europe.newsweek.com/us-calls-americans-leave-gaza-soo...
Electric vehicles with solar charging would be very cost-effective as a border security mechanism.
Where are all the border security startups? Lots of government cash on the table I would think.
A wall, doesn't need electricity.
The maintenance cost is lower.
Can't be hacked.
Oh, and you can paint nice big pictures on it.
Indoctrination [1] doesn't let you question or criticize your programming, unlike modified Asimov's Laws [2].
[1] https://en.wikipedia.org/wiki/Indoctrination
[2] https://en.wikipedia.org/wiki/Three_Laws_of_Robotics