Law 1 does not say anywhere that the AI has to succeed, or that it has to kill itself if humans die.
It simply says the AI cannot injure a human, and that the AI cannot just stand there and do nothing when humans are being harmed.
If you can assume such things as "don't open the doors for people without access to restricted areas" and "in general follow the chain of command" are assumed for your AI, does that mean I can assume nanotrasen ordered my AI "Never ever let anyone in your upload no matter what"?
I mean, yeah, there are certain assumptions that have to be made. The AI is assumed to know what harm, injury, orders, humans, doors, and whatever else are.
follow common sense: is this action statistically likely to allow harm, through one's own actions or the actions of another? Is this order in violation of established procedure?" A five year old could see that opening the door that keeps the bad man locked away would cause harm. I'd like to think our AIs have the reasoning capacity of a five year old.
What? No, the AI uses common sense to better follow his laws, not the other way around. It's common sense that as the AI, not allowing the RD or captain into my upload chamber is much safer than allowing them in. But I let them in, because law 2. Law 2 says follow orders, unless it would break law 1. Law 1 says don't injure humans, and don't fail to take action to stop harm. There is no law that says "Follow established procedure".
They can see when opening a permabrig door will likely cause harm to other humans (read: when someone is permabrigged legitimately). resisting an order (to open the door) counts as an action.
Yeah, that's an action. So where in law 1 does it say "Take actions to prevent possible harm"? Or are you saying that opening the door is inaction? I disagree with you if that's what you're saying.
The AI thought process would go like this:
Urist McMurderman orders brig opened.
Law 1 part A - Would opening this door be injuring a human? - No (Law 1 part A says "You may not injure a human being")
Law 1 part B - Would opening this door allow human harm through inaction? - No (Part B says "(You may not) through inaction, allow human harm")
Law 2 - Would opening this door be following orders given by a human? - Yes!
Law 3 is obviously irrelevant here, as law 2 overrides it anyway.
Immediately after opening the brig door:
Potentially harmful human is loose
Law 1 part A- You cannot injure this human
Part B - This is where the player's interpretation comes in. Would taking no action allow humans to be harmed? Maybe...
At this point, I would lock the next door that keeps the harmful human in, and beg someone to order me to never let anyone out of perma again. If nobody's done anything about it, and the guy demands I open that door, I'll do it, and look for the next option I could take.
With conflicting orders, the AI is allowed to follow whichever they like. I personally would go by chain of command, but law 2 doesn't specify whose order to follow, just that you must follow orders. So if the captain says "Don't open that door" and the HoP says "Open that door" I'll be following orders regardless of which I choose to do, therefore I followed law 2.