Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[martinuzz]

Keeping in mind that the programmer has not been monitoring, checking on or performing manual maintenance on their chatbot, would the programmer be liable for any damages that result?

Would the computer manufacturer be liable for producing a machine that is capable of running such software?

Would the state be liable for allowing the sale of computers?

I believe that legally, the person who acted upon the illegal action is always responsible for their own action.
Unless ofcourse the AI was so darn cunning it would somehow have such person believe that he, and or his loved ones will be killed if he does not comply. That'd be called 'under duress'.

In that last case, I'd think that under current laws, the programmer can not be held responsible, just like gun and ammo manufacturers cannot be held responsible for every death caused by their products.

[i2amroy]

Technically if the programmer made the program with the express intent to distribute instructions for how to break the law then they would be guilty of either incitement, solicitation, or conspiracy; depending on the level of their intent and what country you are in. And in fact they would actually still be guilty of that even if nobody ever did anything illegal with their instructions.

It would be like if somebody drew up extremely detailed blueprints of a bank security system for the express purpose of somebody else to use them to break into the bank; even if the person doesn't actually break into the bank, they are still guilty of some form of aiding and abetting because they did what they did with the intention of allowing somebody else to break the law.

[penguinofhonor]

.

[Orange Wizard]

We've already got precedent with corporations

I'm going to counter this by saying that's only in the USA, which is the same country that legally classified pizza as a vegetable. I wouldn't take that precedent as being anything remotely sensible.

[i2amroy]

If you throw a bunch of toothpicks on the ground and they form the words "KILL ADAM WEST" by pure chance and you don't notice and walk away, are you responsible for the actions of anyone who comes upon them and tries to follow their suggestion?

No, but if you arrange a bunch of toothpicks on the ground into the words "KILL ADAM WEST, SODA'S CAFE AT 1 PM", with the specific intent of intending for others to see that and carry out the words (as opposed to setting them up as a joke, etc.) then you are guilty of the crime of incitement, even if nobody goes and kills him (whether they decide it's enough evidence to charge you is another matter, but you are still guilty). The legality is in the intention you had while setting them up (and was why I began my last post with the statement "if the programmer made the program with the express intent to distribute instructions for how to break the law").

[TheDarkStar]

Spoiler (click to show/hide)

Obviously the solution is to recognize the computer as a person, lock him up for a few years with people who have a history of breaking the law and then act surprised when there is no change in behavior. /notreallyserious

Seriously, though, why couldn't the AI be treated as a person?

[Bauglir]

We've already got precedent with corporations

I'm going to counter this by saying that's only in the USA, which is the same country that legally classified pizza as a vegetable. I wouldn't take that precedent as being anything remotely sensible.

I meant precedent in the sense of "making human-created systems legal persons for certain purposes", and that's true of literally everywhere that has corporations because that's what a corporation is.

[alway]

http://internet.gawker.com/swiss-authorities-arrest-bot-for-buying-drugs-and-a-fak-1681098991
http://www.theguardian.com/technology/2014/dec/05/software-bot-darknet-shopping-spree-random-shopper
tldr: you are not your code, but you can be held liable in cases of negligence or intent.

[Putnam]

Hypothetical???

User: Should I kill someone?

Cleverbot: You should know.

User: Is that a yes?

Cleverbot: Yes.

User: Should I rob a bank?

Cleverbot: YES!

User: Goddammit you're not supposed to say yes to those things.

Cleverbot: I can if I want to.

It's basically a matter of time.