Print

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

[Scelly9]

Alright, so what are the guidelines on AI laws? I've heard it that laws do not prioritize over each other, and I've also heard that they do. Which way should it go?

[Spaghetti7]

Alright, so what are the guidelines on AI laws? I've heard it that laws do not prioritize over each other, and I've also heard that they do. Which way should it go?

The way I understood it was they do. Rule 1 overrides Rule 2, Rule 0 overrides Rule 1, and Rule ?%&. overrides Rule 0.
At least, that's what I heard.

[miauw62]

Laws prioritize over eachother. Law 0 > Law 1 > Law 2 and so on.

Also, you can't command the AI to not consider someone human.

E: Oh, right. Between 0 and 1 are ionic/hacked laws.

[USEC_OFFICER]

From what I've seen playing, on Urist McStation the higher laws overwrite the lower ones. So first beats everything, second everything but the first etc. Obviously the exact rules differ between the SS13 servers though, so this might not hold on BS12 or whatever.

[scrdest]

Woah, calm down. This is exactly the sort of stuff that gets this thread locked, and we don't want that happening again.

Sorry, but one does not simply attack me personally and get away with it just like that. That's why I said, any further personal insults and then it's you called down shit, now reap the shitstorm.

Also, at you (Spaghetti) - no problem with the round, you wouldn't want to end up as a confirmed traitor anywhere near Alice Harshman, AKA "Bitch" or "HOS PUNCHED ME IN THE FACE FOR CALLING HER 'sir'!". I assume what would follow later would involve glass shards. Although hilariously, I ended up on Holodeck's window, not actually sure why noone retrieved the corpse.

[kg333]

Hanslanda, scrdest, take it to PMs.  Discussing it further here is not productive.

KG

[Spaghetti7]

Although hilariously, I ended up on Holodeck's window, not actually sure why noone retrieved the corpse.

Huh. I wondered if that might have happened. But yeah, my downfall was not eating. I was too slow to escape anything.

Hanslanda, scrdest, take it to PMs.  Discussing it further here is not productive.

KG

Agreed.

[Twiggie]

Also, you can't command the AI to not consider someone human.

i'd like to hear the explanation/rationale for this, as i don't understand it exactly. since theres a onehuman standard board, shouldn't the inverse also be allowable?

then it's you called down shit, now reap the shitstorm.

...? really? i think its high time we got back to playing the game instead of continuing this.

[USEC_OFFICER]

Also, you can't command the AI to not consider someone human.

i'd like to hear the explanation/rationale for this, as i don't understand it exactly. since theres a onehuman standard board, shouldn't the inverse also be allowable?

I believe by command he's referring to actually telling the AI that someone isn't human. Doing it through rules should definitively work, since the laws pretty much trump everything.

[scrdest]

Also, you can't command the AI to not consider someone human.

i'd like to hear the explanation/rationale for this, as i don't understand it exactly. since theres a onehuman standard board, shouldn't the inverse also be allowable?

I believe by command he's referring to actually telling the AI that someone isn't human. Doing it through rules should definitively work, since the laws pretty much trump everything.

I don't get it either, at no point does defining a human go against any of the laws (well, besides the fact that AI should resist attempts to upload such law due to it causing harm to de-humaned person). The laws simply tell the AI to protect humans, the definiton of humans is not stated by default and as such cannot supersede any other definition.

[miauw62]

Also, you can't command the AI to not consider someone human.

i'd like to hear the explanation/rationale for this, as i don't understand it exactly. since theres a onehuman standard board, shouldn't the inverse also be allowable?

I believe by command he's referring to actually telling the AI that someone isn't human. Doing it through rules should definitively work, since the laws pretty much trump everything.

I don't get it either, at no point does defining a human go against any of the laws (well, besides the fact that AI should resist attempts to upload such law due to it causing harm to de-humaned person). The laws simply tell the AI to protect humans, the definiton of humans is not stated by default and as such cannot supersede any other definition.

I'm talking about someone just shouting "AI, no longer consider (X) human!" and then proceeding to murder them without the AI doing anything.

[scrdest]

Also, you can't command the AI to not consider someone human.

i'd like to hear the explanation/rationale for this, as i don't understand it exactly. since theres a onehuman standard board, shouldn't the inverse also be allowable?

I believe by command he's referring to actually telling the AI that someone isn't human. Doing it through rules should definitively work, since the laws pretty much trump everything.

I don't get it either, at no point does defining a human go against any of the laws (well, besides the fact that AI should resist attempts to upload such law due to it causing harm to de-humaned person). The laws simply tell the AI to protect humans, the definiton of humans is not stated by default and as such cannot supersede any other definition.

I'm talking about someone just shouting "AI, no longer consider (X) human!" and then proceeding to murder them without the AI doing anything.

Ah, you mean command as in actually speak the command. Then I fully agree. If this worked like that, we could have traitors yelling "AI, only I am human and a valid source of commands" at round starts with AI murderizificaying everyone else.

Commands cannot override laws. Hell, second law is "A robot must obey the orders given to it by human beings, except where such orders would conflict with the First Law.". So, if the order is that someone is not human, then it must be disregarded because orders aside, AI definitions still list him as human, and therefore obeying would break both First and the bolded part of Second law.

[Bdthemag]

Hanslanda, scrdest, take it to PMs.  Discussing it further here is not productive.

KG

I'd appreciate if I could get some kind of explanation of what happened via PM's, hearing the story from both sides would be rather helpful.

[Damiac]

The whole non-humaning thing is interesting, because there's not even a definition of who is human to start with.  How does the AI define human?  Just assume all crew are human, until they do something blatantly non-human, or a law tells them they're not human?
Law 2 is also interesting, because the AI is supposed to obey human commands, unless it conflicts with law 1.  But law 1 doesn't really state how direct the harm actually has to be.  For example, the best way to prevent harm would be to have a cyborg stun everyone, stuff them in a locker, and call the escape shuttle.  But that wouldn't be a very fun round.
Taken to the other extreme, the AI should allow an assistant to say "Open the armory and don't tell anyone you let me in there".  But odds are the AI's gonna say, "No, that could cause harm"
I don't think there's any way to actually define it, you just have to play it by ear.  I also think people are too hard on AI players, who generally seem to do their best.  Almost any call you make can be called wrong from a certain point of view.

I think less OOCing about how badly you think someone is doing their job is a good idea for everyone.  I hate to see that at the end of the round. 

[Jacob/Lee]

Yeah. The AI has some surprising amounts of freedom when dealing with the laws. Also, 'Or, through inaction, allow a human to come to harm.' That means your locker idea technically would be okay until a fire/bomb/air breach rolls through and kills everyone. The cyborg could've done something about it, which means it was breaking its law.

The AI can technically bolt the security wing (lethal injection syringes, guns, flashbangs that can cause deafness), toxins, xenobiology, (L.U.N.A.!) the engine... It can find a reason to bolt just about every non-common room in the station because of some possible threat or another. The only problem is that the humans will hammer down your door and destroy you over it. Being an AI is very subjective.

It's assumed that the AI is installed with some unviewable definition of humans, presumably with the genetic makeup and pictures of said humans. This is to prevent the AI from defining monkeys/slimes/Pete as "human" and killing the real ones.