Refresh the anti-validhunting ruling for asimov ai

[TheRex9001]

The current ruling is: "Asimov silicons that take action against non-harmful human traitors (I.E. those who are committing minor crimes like breaking and entering or theft, or those not carrying or known to possess lethal weaponry) directly and without provocation will be held accountable for any human harm that occurs as a result of their action."
This is kinda bad because it goes contradictory to "Silicons are not Security and do not care about Space Law unless their laws state otherwise" and this ruling leans on crimes. I would suggest changing non-harmful to non-harmful to humans since asimov ais shouldnt really care about non-humans. Traitors should also be changed to "antagonists". Really I would prefer this ruling to state something like "Asimov silicons that take direct action (Like telling security) against human antagonists that are not causing human harm will be held accountable for any human harm that occurs as a result of their action" or similar.

[Timberpoes]

They're complementary.

Silicons don't have to care about Space Law full stop - whether it's a human or a non-human - unless their laws state otherwise.
Asimov silicons that take action against non-harmful human traitors [read: antags] directly and without provocation will be held accountable for any human harm that results.

So as Asimov, don't rat out a human to security unless you have a really, really good reason - otherwise you're responsible for any human harm that results. "They bwoke spess law" is not a good reason.

You can rat out a non-human, but you don't have to and the admin team won't get on your case if you didn't act like Police Constable Borg.

[Vekter]

Yeah, this seems like it's working as intended. It allows for a fair bit of freedom while still limiting AIs to the actual rules.

[MooCow12]

It's kind of paradoxical, you can't assume security is harmful to antagonists and defend those antagonists prematurely, but you can't call out those antagonist's locations so they can more easily/peacefully be detained because that might lead to security harming them.


So when are silicons allowed to predict likely harm from so far nonharmful individuals and when can't they because clearly a line was drawn somewhere with security and antagonists.

[TheRex9001]

It's kind of paradoxical, you can't assume security is harmful to antagonists and defend those antagonists prematurely, but you can't call out those antagonist's locations so they can more easily/peacefully be detained because that might lead to security harming them.


So when are silicons allowed to predict likely harm from so far nonharmful individuals and when can't they because clearly a line was drawn somewhere with security and antagonists.

Its unclear which is why a headmin ruling is good. This ruling needs refreshing at least in my opinion

[Constellado]

Yeah I always get worried as AI when sec asks me to locate a human soon after opening the armory.

I want to follow the order but they look like they are going for lethals...

If I follow the order, I worry I am going to get bwoinked or am breaking the rules somehow? But I also feel that way if I deny it because they usually instantly ask me to state laws...

[Redrover1760]

Yeah I always get worried as AI when sec asks me to locate a human soon after opening the armory.

I want to follow the order but they look like they are going for lethals...

If I follow the order, I worry I am going to get bwoinked or am breaking the rules somehow? But I also feel that way if I deny it because they usually instantly ask me to state laws...

This is a gray zone. Either option works and neither will get you bwoinked. You have probable cause to refuse but dont have to.