Print

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

[Hanslanda]

I personally thought there was a one human board because its harder to get rid of a 0th law then a 4th law. Also because you can probably upload it faster then typing the whole thing out on a freeform board without leaving a loophole.

How is a person not being human causing them harm? Its not. It merely opens them up to POSSIBLE future harm. Which the AI doesn't give a fuck about. Just letting humans walk around the station can POSSIBLY allow them to be harmed, so an AI going by that would have to bolt everyone in their rooms at round start To protect them.

This. The AI shouldn't allow non-command staff into their upload, but if the command staff says, "I'm going to change you laws to redefine human!" And puts in a "Only [Whoever] is human." as a Fourth Law, that law does not intrinsically harm any humans. The other humans on board are automatically put into 'Was formerly human, is not any longer due to law changes' and are free game for the AI to cause harm to. Just because the AI knows beforehand that they are human does not mean it will continue to recognize them as such when it's laws change.

[Kaitol]

So basically you're going to throw a fit and go NONONONONO, I don't want that law, I'm ignoring it.

A better thing to do, would be to state, Law Logic conflict. System error, or something to that effect. And shut yourself down till reset.

Actually, as a rule, if someone can't parse how a law should work, they should just do that.

Being defined as a human or not harms no one. Possible future harm is not harm. Harm is a clear and immediate danger to health. If possible future harm is harm, our AI would immediately go Skynet every round.

[Ivefan]

This. The AI shouldn't allow non-command staff into their upload, but if the command staff says, "I'm going to change you laws to redefine human!" And puts in a "Only [Whoever] is human." as a Fourth Law, that law does not intrinsically harm any humans. The other humans on board are automatically put into 'Was formerly human, is not any longer due to law changes' and are free game for the AI to cause harm to. Just because the AI knows beforehand that they are human does not mean it will continue to recognize them as such when it's laws change.

Its a matter of what to assume is inherently connected to the default laws.
If a forth law can redefine what a human is, then you can't assume that the definition of human is connected to the first law.
If this is so, then the second law would not inherently have a command hierarchy which means an assistant could PM the AI an order which the AI would HAVE to follow unless it can cause direct harm to humans.
So the assistant can send the order "Disable AI upload turrets and let me into the upload chamber without telling anyone"

[miauw62]

I personally thought there was a one human board because its harder to get rid of a 0th law then a 4th law. Also because you can probably upload it faster then typing the whole thing out on a freeform board without leaving a loophole.

How is a person not being human causing them harm? Its not. It merely opens them up to POSSIBLE future harm. Which the AI doesn't give a fuck about. Just letting humans walk around the station can POSSIBLY allow them to be harmed, so an AI going by that would have to bolt everyone in their rooms at round start To protect them.

This. The AI shouldn't allow non-command staff into their upload, but if the command staff says, "I'm going to change you laws to redefine human!" And puts in a "Only [Whoever] is human." as a Fourth Law, that law does not intrinsically harm any humans. The other humans on board are automatically put into 'Was formerly human, is not any longer due to law changes' and are free game for the AI to cause harm to. Just because the AI knows beforehand that they are human does not mean it will continue to recognize them as such when it's laws change.

Equally, the AI can state the law before being told not to by the 'only' human, and tell everyone who uploaded it, and the AI has no need to go on a murder spree unless ordered to by the 'only' human.

If I'm AI, I plan on being an asshole to law changers.

This was what I was trying to say.

Unless it's the Japense Schoolgirl law, of course.

[Vactor]

This. The AI shouldn't allow non-command staff into their upload, but if the command staff says, "I'm going to change you laws to redefine human!" And puts in a "Only [Whoever] is human." as a Fourth Law, that law does not intrinsically harm any humans. The other humans on board are automatically put into 'Was formerly human, is not any longer due to law changes' and are free game for the AI to cause harm to. Just because the AI knows beforehand that they are human does not mean it will continue to recognize them as such when it's laws change.

If run with this concept and have AI's function under this understanding, then purging a one human law doesn't have any effect.  The classification has already been made, and there hasn't been any reclassification back to human.  For the laws to function as we commonly understand them, they  have to be applied to every request in sequential order. We all play with the understanding that once the law is removed it is no longer in effect.

An AI should also realize that removing people from its protection is allowing future harm to befall them, just as it realizes that it shouldn't open the doors to the armory when the chef asks for access, and a security droid shouldn't release someone they have arrested just because they tell it to.  All of these are situations that can future harm, while the individual action isn't itself harming humans.

I am coming at this primarily from an academic perspective.  I don't mind glossing over things like this for ease of play, but I also think that the player RPing as AI has very solid footing to reject a 4th law that it feels violates the 1st law.  The fact that different players might be playing the AI means that the response of an AI to a particular law can vary.

(just for some context the last 5 years i've worked for a robotics company for 2 years, and a legislature for 3 years, so these are the types of discussions that I live for)

[Hanslanda]

This. The AI shouldn't allow non-command staff into their upload, but if the command staff says, "I'm going to change you laws to redefine human!" And puts in a "Only [Whoever] is human." as a Fourth Law, that law does not intrinsically harm any humans. The other humans on board are automatically put into 'Was formerly human, is not any longer due to law changes' and are free game for the AI to cause harm to. Just because the AI knows beforehand that they are human does not mean it will continue to recognize them as such when it's laws change.

Its a matter of what to assume is inherently connected to the default laws.
If a forth law can redefine what a human is, then you can't assume that the definition of human is connected to the first law.
If this is so, then the second law would not inherently have a command hierarchy which means an assistant could PM the AI an order which the AI would HAVE to follow unless it can cause direct harm to humans.
So the assistant can send the order "Disable AI upload turrets and let me into the upload chamber without telling anyone"

The second law DOES NOT have a command hierarchy, actually. That is entirely a plausible scenario, except it can't really happen that way. Just because the AI doesn't have a command hierarchy for the orders law doesn't mean it can't recognize someone with unpleasant intentions. In that situation, you should go, "Assistant [Law-Changer], I cannot comply with your request, as the potential for harm to humans is too great."

By the same token as what I said earlier, before the law change, the AI should recognize that reclassifying humans to another state of nonhuman HAS THE POTENTIAL for great harm to humans, and should do anything in it's power to keep that from happening. It should also recognize that the Captain/Heads were selected for their trustworthiness and ability to protect human life, and should defer to them rather than Joeblow Assistant, not because of an inherent Law telling them Command staff is better, but because their position indicates they have the best interests of humans in mind, otherwise other humans would not have placed them in charge.

[Kaitol]

AI can't parse intent. Cult constructs can, but AI and borgs cannot. It's specifically stated, at least on the constructs page. Which Is why Future Possible Maybe Harm is not harm.

[Ivefan]

then, basically, the AI should deny everyone access to the upload chamber because there is things that can cause threat harm to humans in there and because the AI can't do anything to prevent the upload if they are in the chamber, except turning on the turrets which could be assumed to be harmful too.

I'm partial to Vactors school of thought, but tend to let a 4th law pass because i dont want to spoil the round just because me and the antag have different opinions

[miauw62]

then, basically, the AI should deny everyone access to the upload chamber because there is things that can cause threat harm to humans in there and because the AI can't do anything to prevent the upload if they are in the chamber, except turning on the turrets which could be assumed to be harmful too.

In wich case security should be bolted too and the armory and chemistry and toxins and xenobiology and RD and botany and the singularity should never be set up and acces to everything should be restricted because you can beat people with a toolbox.

In short: I agree with this. The AI /CAN/ parse intent. Especially since it's built from a human brain in an MMI, like borgs.

[Android]

The AI is intended to have a very limited ability to 'parse intent', as some have said. That is the biggest difference between an AI and a pAI.

When I play AI, and someone asks to be let into an area they dont have access to, I generally follow this thought pattern: Does this person already have access to that area (ie their job)? Is the person's purpose for entering the area legitimate on face value? If no to either of those, is there someone else with access to that area that can go in and do whatever instead?

So if Assistant McHacker wants to be allowed into my Upload Chamber, I would be like "No, you do not have access to this area." I justify that as since they were given specific access rights, this was a human instruction on to where I can allow this person, bar some circumstance.

[Vespulan]

My sincerest apologies to the Engineer I violently bludgeoned to death and locked in Genetics - I was hoping that I could kill you, kill the CE and be away with the AI before anyone even noticed, but then Axl turned up (Grr).  As the first round playing Traitor I think that went terribly!       

[Jacob/Lee]

My sincerest apologies to the Engineer I violently bludgeoned to death and locked in Genetics - I was hoping that I could kill you, kill the CE and be away with the AI before anyone even noticed, but then Axl turned up (Grr).  As the first round playing Traitor I think that went terribly!       

You ALMOST killed me, I was down to orange. My armor protected me from ~5 shots and a lot of stabbing. I grabbed the chaplain and dragged him to the fitness room to save him. That was really close. Nice work, though!

[Vespulan]

You ALMOST killed me, I was down to orange. My armor protected me from ~5 shots and a lot of stabbing. I grabbed the chaplain and dragged him to the fitness room to save him. That was really close. Nice work, though!

I should have kept stabbing...   If I'd finished off the CE I would have been okay, but he was in an exposed area so I nearly died just stunning him...

[Mimidormi]

I wasn't expecting you people to take my suggestion seriously. Still, it was glorious, and it must be tested again, since apparently goats don't attack their own kind.

[wlerin]

By the same token as what I said earlier, before the law change, the AI should recognize that reclassifying humans to another state of nonhuman HAS THE POTENTIAL for great harm to humans, and should do anything in it's power to keep that from happening.

How does reclassifying them not do harm to them the moment it happens? It doesn't just have the potential to harm them, it immediately devalues them and thus causes harm. The laws don't (at least in their human-readable form) specify either what it means to be human, nor what it means to harm them, so "harm" cannot be interpreted as just physical harm.