An interesting ethical question(s)...

It is 2014. An intelligent, autonomous drone is flying a reconnaissance mission over Iran when it is illuminated by an Iranian air-defence radar. It defends itself by firing a missile, which hits the radar and kills eight radar operators.

Question: Should humans EVER build autonomous machines with the right to self defence that may involve killing humans? How does that square with Issac Asimov's three rules for robots?


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 to 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.

humanity demands the right to fuck shit up

to answer your question: one would think building such machines wrong, but I am sure in the future we we do for very good reasons - so we can invade the planet vads and put to death their sentient robot Kantotrons

The day it can accurately judge whether the destroying of the missile site would also entail the possible harm to humans yes. Otherwise, no.

We are decades - if not longer - from having to worry about it in reality. And by then technology to track the whereabouts of every human on the planet might exist or, more likely, have the ability to disable a missile without blowing it up (EMP or other type of smart weaponry).

Depends if iran and the drone owner are at war and whether there is a *possibility* that it might respond to civilian 'threats'.

To be used in war it should meet some standard of risk that is less than that of humans doing the same job (cf fighter pilots/ friendly fire).

On the theme of the original post - machines, dumb or otherwise can easily kill humans by mistake, equipping them with narrow impact weapons is not a huge step up from where we are now.

It comes down to the situation that machine operates in.

An example would be mine/claymore or perhaps an auto sentry gun, all of which are in use today.

Really? It was this article that got me thinking -

http://www.thenewatlantis.com/publications/military-robots-and-the-laws-of-war

"...When U.S. forces went into Iraq, the original invasion had no robotic systems on the ground. By the end of 2004, there were 150 robots on the ground in Iraq; a year later there were 2,400; by the end of 2008, there were about 12,000 robots of nearly two dozen varieties operating on the ground in Iraq... ...just as any human’s right to self-defense is limited, so too should be a robot’s. This sounds simple enough, but oddly the Pentagon has already pushed the legal interpretation that our drones have an inherent right to self-defense, including even to preemptively fire on potential threats, such as an anti-aircraft radar system that lights them up. There is a logic to this argument, but it leads down a very dark pathway; self-defense must not be permitted to trump other relevant ethical concerns...

If you are at war then why is it a concern if it fires at a radar system? If you aren't at war then 'self defence' would fix that quickly.

Because in my initial question I said it was an AUTONOMOUS robot... Is it ever ethical to arm and allow autonomous engagments by such machines is the question.

An interesting question - if the human commanders have programmed clear rules of engagement without "wiggle-room" in the same way they would have for a human pilot then it could perhaps be excused.
i.e. If human pilots were told unequivocally to take out radars that light them up on a given mission, then giving the same orders to an unmanned plane sees no difference in outcome.

there is one difference - can robots be put on trial for war crimes?

cause pilots and commanders most certainly can

would the 'just following orders' defence actually work for these machines

*vows not to go to the forthcoming terminator movie*

Possibly,

If you are at war with whoever is in the area in which it is operating.

And

It can distinguish between military attack and civilian investigation

then perhaps it is no worse than what we have now.

The current humans dont seem to worry about civilians when they are near military targets so its hard to be any worse.

It's only an extension of what currently exists.

The ALARM missile IIRC will launch at a known radar source or even an area of interest. It'll then deploy a parachute and float down. If an illuminator then lights up, it'll attack that illuminator.

There are other missiles that are fired into an area and then use their own active seeker to acquire a target (the link between operator and weapon is broken)

A Phalanx CIWS, once turned to active or whatever, will engage anything that approaches it that fits certain parameters. It is entirely autonomous during that period. I believe that the Aegis system can also be set to an autonomous mode as well.

Then there are encapsulated torpedoes - mines that sit on the surface and activate when a target goes past.

How are any of them really any different? We've had weapons that don't require point-and-shoot for 30+ years.

according to stratfor its already happening

this from an email y'day re recent terrorist attacks in Pakistan

" In his claim of responsibility, Mehsud said the Manawan attack was in direct response to the expanding U.S. unmanned aerial vehicle (UAV) campaign in Pakistan. Mehsud threatened that there would be more militant attacks in Pakistan and the United States if the UAV attacks did not stop"

lol yeah they were holding stuff in reserve before this.

They are controlled from the ground so they may aswell have said we will continue the attacks while the US exists.