It looks like professor
Noel Sharkey isn't the only one stirred up about
autonomous killing machines. An anti-landmine group in London called Landmine Action is now calling for the ban of robots capable of killing people all on their lonsome. The group believes these bots fall under the same category as landmines -- which are outlawed in 150 countries -- and is pressing governments to keep control in the hands of a human operator. The group is also campaigning to put a stop to autonomous cluster bombs (which self-detonate), and feels these bots fall under the same category. Says spokesman Richard Moyes, "Our concern is that humans, not sensors, should make targeting decisions. So similarly, we don't want to move towards robots that make decisions about combatants and noncombatants." Of course, if the bots can
pass a Turing test, how will we know the difference? Needless to say, Skynet could not be reached for comment.
posted March 28th 2008 11:07AM
I'm pretty sure being a suicide bomber is already illegal.
Oh yea? Then why isn't anyone ever prosecuted for it?
I agree with this...mostly. However, I think that its morally ok for machine AI on the battle field to do its own targeting, so long as a human operator is the one that gives it the green light after the fact to engage the target. A human operator should be in charge of high level decisions, not the low level ones. So essentially, the robot can do everything except a human has access to a master safety on the weapon.
remember folks, zap brannagin can always send wave after wave of his own men at them!
"a sad day for robot kind... well, we can always build more killbots!"
-bender
Autonomous robots are totally different than landmines because you can turn them off. Or else we should outlast the conventional bomb in wars as well because that kills just indiscriminantly. ie. how is a bomb with a 1/4" mile detonation radius different than putting an autonomous robot and programming it to shoot anything within a 1/4" radius for a set amount of time? Or when we use laser guided missiles to blow up a building?
If anything even with a flawed combatant recognition, autonomous robots would still be more civilian friendly than any non-hand carried weapon.
Dear Engadget,
Please use a picture of Johnny 5 in ever post. Johnny 5 loves the EEE PC! Johnny 5 doesn't like the Kindle! Johnny 5 loves the iPhone! etc, etc.
NO DISASSEMBLE STEPHANIE!
EMP grenades anyone?
Wow...
Repetitive writing FTL
Shouldn't "KILLING" in general be outlawed?
Resistance is futile, you must comply.
What are you talking about? Killing is for the most part outlawed. Can you give a better definition of what you mean by "outlawed". Do you mean that instruments of killing should be outlawed? Just explain how you can outlaw instruments of killing without using instruments of killing and how you can "outlaw" conflict without resorting to conflict? Its a big catch 22.
I think robots will make better judgements than we do, it's just a matter of time, their sensors and processing have the potential to be much faster and more acurate than ours.
Robots SHOULD be capable of making better decisions of who to kill and who not to kill based on a variety of sensors.
They should be able to determine the THREAT LEVEL of an individual. Maybe that entails whether or not he/she has a gun or hidden explosives (x-rays, microwave sensors,etc).
Perhaps they can discern vocalized pleas for help - or threats with a microphone.
Perhaps they have thermal imagining to determine the person's anxiety level.
One thing I do like about a robot killing machine over a Human Killing Machine is that robots AREN'R RACISTS. They will shoot you dead, but they'd have to be programmed to use racial slurs. They won't discriminate on people based on how they "look" or how they smell. They won't kill people because they are a different color, or have a different sexual orientation (unless programmed to do so).
A machine acts just as it's programmed. You say they're not racists? But what if a profiling system that includes race is a part of their threat assessment. Somebody like Greenteacookiedough could be a part of the development team. Being a machine doesn't keep the programmers from including bad evaluation mechanisms.
are we sure this wasn't built just to lay massive amounts of pwn in BattleBots?
http://www.engadget.com/2008/03/14/battlebots-headed-to-espn-near-you-now-with-less-suck/
That pic of Johnny 5 never gets old. I'm really suprised they didn't use one of him driving the lunch wagon for the Darpa vehicle articles.
Didn't read all the posts so lower me b/c I double posted, but a robot isn't going to pass the Turing Test anytime soon, if at all. Most humans can't pass the Turing Test ;)
I don't like the idea of this. Not because of the robots' accuracy or anything like that, but rather how unethical it is. High-tech pieces of machinery should never be allowed to take the life of a human away.
your mama was a snow blower! Lol damn, #2 said it, good call!
I'm just reading this article for the great Johny 5 references! XD
this issue reminds me of the opening scene in robocop. maybe it would be best to leave targetting decisions in the hands of humanoids.
Holy crap its Johnny 5, please don't bleed like a human johnny, that sh*ts just too real for me