[Vekter] Digresskew - Manuel AI warning for spacelaw

[Digresskew]

BYOND account: Digresskew
Character name: Cargo Cult
Ban type: Red Warning
Ban length: N/A
Ban reason: "Manuel - As AI, spent a solid 20 minutes assisting Sec in arresting the CMO for breaking into the Captain's quarters despite multiple Law 2 requests to do otherwise. Also spammed vox announcements a bit, don't do that."
Time ban was placed: 2020-02-04 15:35:55
Server you were playing on when banned: Manuel
Round ID in which ban was placed: Cannot be found, there is no ban page because this is a warning.
Your side of the story: As an AI, I had assisted the Chief Medical Officer at the time in entering the captain's quarters. Although I did not ask them at the time, I had assumed they were simply attempting to secure the disk. However, upon surreptitiously declaring themselves acting captain, they then proceeded to grab the captain's locker and try to run out with it. (without securing the disk, I might add) As there was a Head of Personnel at the time, and a research director who was arguably of equal rank and unconsulted on the matter, AND the CMO had not been declared captain in any official or unofficial capacity aside from their own declaration, I viewed this as a capital offense under space law. Because two of the possible punishments under space law are death, and thus are apparently harmful, I had decided to impede their attempt under law one. I had told them that they were entirely free to leave the captain's quarters, so long as they left the locker behind. After a long time of back and forth between me and the CMO over the radio, publicly, I eventually asked security to apprehend them before they did anything that would ACTUALLY get them harmed. the Captain's locker was subsequently stowed in the armory.
Why you think you should be unbanned: I believe that this note is unnecessarily restrictive on AI behavior and interpretation of Asimov laws. Even for Manuel, I think that inhibiting capital offences from proceeding is an entirely justified action according to law one. Vekter clearly thinks differently. I wanted to introduce this precedent to other admins and see what they thought.
References of good conduct: N/A

[Vekter]

I think that you're misunderstanding the purpose of the note. I wasn't advising you not to step in, I was advising you not to spend a solid twenty minutes hunting the person down and playing security AI because they broke into the captain's quarters for their locker. You were asked multiple times per law 2 by the CMO to allow them to leave when they posed no clear and present danger to human lives and ignored it under the basis that they might be executed. You're allowed to prevent people from being executed, but this just felt excessive. You had no reason to believe that the other heads would have actually killed them.

The AI doesn't really need to interpret space law in any way, shape, or form. You have one purpose on the station - assist the crew while holding to your laws. None of your laws stated anything to do with space law.

As usual, other admins are free to chime in, but I'm leaving this placed for now.

[Not-Dorsidarf]

Just as a side notes, all heads are equally entitled to be acting captain as per server standards, if the HOP doesn't promote himself with his first dibs at the post it's fair game.

[SkeletalElite]

Silicons may choose whether to follow or enforce Space Law from moment to moment unless on a relevant lawset and/or given relevant orders.

I think this may be the source of confusion on the appealers part, as per the "given relevant orders" part, law 2 comes before space law.

[Vekter]

Silicons may choose whether to follow or enforce Space Law from moment to moment unless on a relevant lawset and/or given relevant orders.

I think this may be the source of confusion on the appealers part, as per the "given relevant orders" part, law 2 comes before space law.

Peanut, but quoting in case it gets deleted - this is a good point. Your interpretation of space law is trumped by your AI laws.

[Anonmare]

Under silicon policy "Asimov and Human Harm" section 1.1 (paraphrased):-
Humans are allowed to order the AI to let them perform actions that are harmful to themselves if they are fully cogniziant of the consequences of doing so. The CMO should be assumed to know that they were potentially facing consequences and, at most, could have been warned about such actions before their orders were carried out.

[Karp]

The note in question doesn't mischaracterize your situation. You can't assume an action is harmful or will be harmful unless it will cause harm immediately afterwards, AKA imminent harm(Letting a simplemob into a room to kill someone, opening a door to expose burning plasma into another room)

Heads can be the acting captain at any point if no other head takes the position and them wishing to be released isn't an unreasonable request.

The CMO was doing something that was expected and completely reasonable of them, the lockdown itself and calling security is a violation of your laws(Which comes first in interpretation, you must listen to your laws first and then you can listen to any personal creeds/beliefs you want like space law or any gimmicks gimmick.)

These restrictions partially exist because the AI being able to freely interpret their laws to their benefit would be an active detriment to the flow of the game and would make them a near-unstoppable situation ruiner. The note exists to state that you should know better in the future if the same situation ever comes up again. Denied.