Military robots and the future of war [TED] | Hacker News Reader

Military robots and the future of war [TED]

ted.com

19 pointskvh32 comments

He makes some good points, and it's interesting that soldiers who operate drones remotely (from the US) suffer higher rates of PTSD than soldiers actually in Iraq...but I felt like there was also a bit of dramatization, perhaps because he doesn't (seem to) understand robotics/AI very well. For example, he says that robots see an eighty year old woman in a wheelchair the same way they see a T80 tank - just "zeroes and ones". But that's like saying that humans see them the same way - just stimulus on rods and cones. I cringed at that.

Do you see his point if you substitute 'digital blob' for 'ones and zeros'? Sure we see 'blobs' also, but we see the 'meaning' of the granny just about inextricably too; a computer won't.

No, substituting "digital blob" doesn't help at all. Have you ever heard of computer vision? I don't see any reason why a computer shouldn't be able to distinguish a tank from an old woman.

How about, '/Things like/: old women, dogs, decoy tanks, captured tanks, someone cutting up a dead tank for scrap..'.

There may be a generation or two of robot that is deployable in some battlefield scenario in some autonomous capacity, but can not make that determination as well as a human. That will, however, be an anomaly that is rapidly corrected, followed immediately by it making the determination better than a human.

(It should be pointed out that we are hardly perfect at recognition tasks. When discussing robot/computer vs. human capabilities, people often slip in the assumption that humans are perfect at the task in question, usually unconsciously. That is never true. Humans have error rates too.)

Then "some" soldiers actually in Iraq. And I cringed there too. The guy seemed to have gathered a decent amount of information, it's too bad he didn't organize it a bit better.

What concerns me is the completely vacuous presentation, complete with mention of Zune, and hopeful talk that they won't commit war crimes, and maybe humans are the problem. If this is the quality of philosophical thinking going in to this problem, Nazism isn't far away.

Huh? I had issues with the talk as well, but I don't see how anything he said leads to Nazism.

Lack of critical thinking (eg, completely automated rules of engagement algorithms have been deployed for decades). Military worship (toned down, but present).

Ooookay. Those might have been present in Nazi Germany, but I wouldn't say that they're either sufficient or necessary for Nazism.

This was a horrible talk.

Some points were just false (war tech is not open source, china & india don't make US bots), lots of the visuals were misleading.

Some of his points were overly emotional and would better fit a middle school debate than TED.

Not to mention that the "72 virgins" joke [1] was distasteful beyond belief.

[1] I assume it was meant as a joke because I find it hard to believe that someone who gives a talk at TED could be that ignorant.

If it was a joke, it flew over most everyone's heads, especially considering the way the presenter said it.

Ignorant of what?

Of the reality?

Unfortunately, religious fanaticism is a reality. People do die and kill in name of a religion.

That's quite true. But it's also true that the "72 virgins" are not an explanation. The reality is much more complex than that.

The 72 virgins joke was a metaphor for all of the preposterous religious motivations behind suicide bombings.

Although it is true that there are almost always secular motivations as well, the metaphor is not dishonest and the joke is funny.

If you think that all suicide bombers are motivated by religious fanaticism, you really need to read a History book. Once again, that's a simplistic explanation that fails to take into account the full complexity of the reality.

Keep downvoting. It only makes you look like ignorants.

I don't downvote, but really, you should take a hint here. Write a post explaining _why_, and upvotes will come. Simple opinions are not encouraged on HN, they're usually considered spam.

Examples of suicide attacks which had no religious motivation:

- the Luftwaffe carried out suicide missions against the Red Army during the Battle for Berlin.

- the Japanese Kamikaze inflicted great damage in the U.S. Navy fleet in the Pacific.

- the Vietminh / Vietcong who used suicide bombers against the French Military and then against the U.S. Military.

Examples are abundant. It's true that many suicide bombers are motivated by religious fanaticism, but if we look further back in History, the pattern seems to be that suicide bombers are motivated by other kinds of fanaticism (e.g., nationalism) or simply by desperation. When one can't defeat a stronger enemy in conventional ways, suicide attacks can be a good option, as they inflict tremendous amounts of damage with little resources, and have devastating effects on the enemy's morale.

To cut a long story short: not all suicide bombers are created equal. The ones who carried out the 9/11 attacks are quite different from the palestinians who blow up a bus in Tel-Aviv, for instance. Failing to understand this reality is a serious mistake.

Kamikaze means divine wind. They were dying for the emperor who ruled by divine right. There was more religion in that fanaticism than you might think.

You have a point there. But it's also true that many Kamikaze believed that their suicide missions were the only way to save Japan, and hence their families back home.

Open Source projects are used in war technology. Linux powers a lot of robots.

Aren't all of Pentagon's cool toys programmed in ADA?

I never heard of an open-source project based on ADA...

His point was that anyone could do it because it was open source. Robots indeed almost exclusively build in Linux. That doesn't mean their designs are open.

An exception might be the vibrant hobbyist small uav builder.

A hobbyist small UAV builder can indeed cause a lot of damage if he has evil intent. For instance, a small UAV with half a pound of high-explosive could glide over a "secure area" and try to hit a high-value civilian target. Much easier than detonating a car-bomb or trying to take someone out with a sniper rifle. Sharpshooters will have to scan the skies too (if they're not doing it already).

Fortunately, it's not exactly easy to buy IMU's, otherwise we would have heard of guided-missile hobbyists doing evil things. Also fortunately, GPS shuts down above a certain altitude and speed (not applicable to GPS units belonging to the U.S. military).

There are so many more people that know how to drive a car that this kind of threat is still nothing to worry about.

Lots of car bombers don't know there is a bomb in the car before it is remotely detonated.

Also, snipers are really, really effective, and again, are much more common.

Snipers are effective, but they can be detected using laser technology, and they can be neutralized by sharpshooters. What kind of counter-measure could one use against a UAV-bomb?

Watch "A Taste of Armageddon"

http://www.cbs.com/classics/star_trek/video/video.php?cid=61...

http://www.startrek.com/startrek/view/series/TOS/episode/687...

Many months ago when I first learned that USAF and RAF pilots in Las Vegas, NV were flying armed drones in a war zone "worlds" away I knew we were not far from the scenario presented in that episode.

War needs to be painful, it needs to be bloody and costly otherwise there is no reason to stop fighting and no arguments against starting wars and "nuke from orbit" becomes the 1st/easiest/best answer to every problem.

The historical correlation between the in-your-faceness of war, and the amount of war, is negative. (The news fools you; on a per-capita basis, the amount of war is going down. http://www.ted.com/index.php/talks/steven_pinker_on_the_myth...)

There is no actual evidence that making war bloody and costly is an effective strategy for preventing it.

Just the title statement iteself...

I always felt... why freaken robots, just have the leaders challange each other in a game of chess (Kasparov will be king) and basically we make rules based on technological levels, population, and other things, how many advisors you are allowed to have.

They just duke it out and after the one game war is over. Everyone accepts the outcome and the new overlords come to claim the country.

Its basically the same shit as war, except less death. We just need to figure out a way to incorporate nukes and suicide bombing into this game and a way to ensure propper resources are spent. After that why bother sending armies... Just send the chess team.

Reply on news.ycombinator.com