are the three laws of robotics really safe???

[ster_tet]

are the three laws of robotics really safe???

   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.
   
 

[airman00]

are the three laws of robotics really safe???

   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.
   
 

there is not one robot today that can comprehend what any of those things mean . Now if we werer to theoretically say that robots did follow these laws , we would see that there are some occasions where it isn't safe ( read I Robot to see )

[ster_tet]

in what case that it wouldn't be safe?

In the I Robot movie what is the part there that implies the the 3 laws are not totally safe?[/size]

[airman00]

There is a book you know , which came way before the movie
http://en.wikipedia.org/wiki/I,_Robot


The book has very little to do with the movie

[pomprocker]

wasn't there a zeroth law added

[AndrewM]

The first problem with the robotic laws is they can allow a robot to unknowingly harm a human.  The second problem, and where the first problem actual stems from is that they are absolute.  This creates a conflict in what we humans would call morality, and general common sense.

I think one of the examples in the books was having robot A poison a glass of milk while being told the milk was being disposed of.  Robot B is then told to deliver the milk to a person, without ever being told the milk was poisoned.  If people were to replace the robots, it probably would not result in a person being poisoned as common sense would kick in with us thinking "Hey poisoning the milk is probably a bad idea".

Add to that the problem of morals and human (ir)rational thinking.  In the instance of two people in equal mortal danger, only one of which the robot can prevent causing harm to.  A strictly logic conflict with the absolute laws would prevent the robot from taking any action (it goes into a paradox loop logically). 

As an example, consider a robot operating a printing press.  One of the human workers on the printing press has had his shirt grabbed by the press and is being pulled into the press.  The added resistance to the press has caused a jam further along the press resulting in one of the rollers unbuckling and on the verge of falling (very heavy rollers if you have never seen a large printing press) ontop of another human.  If the robot stops the press to prevent person A from being completely pulled in an crushed by the press, they will not have time to stop the roller from falling ontop of person B.  Saving person B means letting the press run for another 10 seconds, likely killing person A.  What does the robot do?

[AndrewM]

wasn't there a zeroth law added

The zeroth law is what can potential prevent paradoxical loops, as well as allowing a robot to harm a person.   It is the "for the good of humanity" law, meaning a robot can allow a person to be harmed, or harm the person directly as long as it determines that doing so is for the greater good of humanity.  So in my above paradox example, person B would be saved while person A gets pulled in the press because the robot determines that person B is a better, more valuable, employee and for the good of humanity must allow person A to die (to save person B).

EDIT:  Just to be clear, it is a very bad law to add because killing off one human at a time to prevent global warming caused by the carbon dioxide produced by that one human (and waste products, and resources used, etc) is really for the good of humanity.  From a logical point of view at least.

[paulstreats]

Are there any real laws? (apart from health and safety, consumer safety etc..) apart from in the movies?

Maybe intelligent robots could just follow the law of the country that they are in. That people follow themselves. There will likely be mistakes but then again lifes are endangered by humans making mistakes or bad judgements - It would just be media sensationalism if a robot killed somebody by mistake

[Admin]

The whole point of Asimov's book is to show where the laws fail

See this:
http://www.economist.com/displaystory.cfm?story_id=7001829

Are there any real laws?

yeap . . . if your robot injures or kills a human, expect your company to get sued out of existence

unless of course its a military robot, in which case the 3 laws are counter-productive . . .

It would just be media sensationalism if a robot killed somebody by mistake

Actually, robots have killed people already . . .
http://en.wikipedia.org/wiki/Robert_Williams_(robot_fatality)
http://thestar.com.my/news/story.asp?file=/2006/1/8/nation/13054882&sec=nation

[paulstreats]

The whole point of Asimov's book is to show where the laws fail

See this:
http://www.economist.com/displaystory.cfm?story_id=7001829

Are there any real laws?

yeap . . . if your robot injures or kills a human, expect your company to get sued out of existence

unless of course its a military robot, in which case the 3 laws are counter-productive . . .

It would just be media sensationalism if a robot killed somebody by mistake

Actually, robots have killed people already . . .
http://en.wikipedia.org/wiki/Robert_Williams_(robot_fatality)
http://thestar.com.my/news/story.asp?file=/2006/1/8/nation/13054882&sec=nation

I did mention the real laws being apart from consumer safety laws (an electric shock from a kettle would result in your company being sued out of existence).

Also the japanese killed by the robotic arm is really just a machine.
My definition of laws regarding robotics would be more like something that must be built into an AI system. Many people have died because of being fed through a steam press machine that can be seen arguably as robotic as an arm that paints / builds things, its usually human error that is to blame as in not enabling safety shut downs, not removing safety guards etc... But has anywhere produced any laws specifically regarding robotics and AI? Im just interested if anywhere is forward thinking enough just yet

[izua]

I don't believe anyone will make laws regarding AI unless that is marketable and used on large scale industry. Industry needs work, AI thinks. Someone who thinks doesn't work said a great thinker once

[Admin]

Someone who thinks doesn't work said a great thinker once

I think a neuron died while trying to parse this sentence, lol . . .

[Private Reid]

  Okkkk...

You guys are going to give me nightmares...






Lol, But still, it's a little     creepy    Makes me look at my robot in a different way...

I'll have to give it a lecure. "Now, I know youve read those internet articles but, i swear to god if you try to kill me ill rip that little head of your neck!!! Burn your microcontroller! Understand!!!"