The coffin of borgy and pris pris nut

[MooCow12]

viewtopic.php?f=7&t=35366

I thought silicons arnt allowed to take security/captain's accusations as gospel like how an AI cant treat someone as a changeling without seeing them use their powers just like an AI cant treat someone as a human harmer without seeing it.

Does this mean AI players get to help valid hunt anyone the captain says is a ling?




Also i think there was left out context, Bible should really argue that the borg's own ai was telling him not to release the prisoner more since that seems to be a stronger arguement. And the only mention of it is in the note.

[Vekter]

I SAW WHAT YOU DID.

Anyway no, that's a different situation because it's the difference between "taking an action that can directly harm a human" (attacking someone the captain says is a changeling) and "not taking an action that could lead to human harm" (not releasing a prisoner). Asimov silicons can't do anything that could lead to harming a human, and if they took someone's word for someone being a changeling and harmed them, they would be violating law 1 if that person wasn't a changeling.

On the opposite side of things, not letting out a potentially harmful prisoner if ordered to do so would be a law 2 violation, but letting him out if he's harmful would violate law 1. Taking the action (letting him out) would lead to potential future harm, so not taking the action when there's any doubt is the correct play.

[MooCow12]

I SAW WHAT YOU DID.

I dont normally make threads i just wanted to make this one for the title :c

[Vekter]

I SAW WHAT YOU DID.

I dont normally make threads i just wanted to make this one for the title :c

Ping me next time and I'll just move it.

[EmpressMaia]

perma ban OP for making that refrence

[MooCow12]

Apparently the ai telling the borg not to release the prisoner happened after the incident so that was the missing context.

[TypicalRig]

Silicons can't assume a person is harmful to deny a law two order unless there's some basis for this assumption. Just being told by another party would fit this bill, but them also not blindly believing the cap without concrete proof is also valid play.

Since the AI didn't counter order it until after, it's a bit odd, but assuming it followed orders afterwards this is veering in grey IC issue territory. But their "weird OOC conversation" sounds like important context so who knows.

[CPTANT]

Warned about potentially releasing harmful prisoners as an Asimov AI - in this case, the prisoner had been proven harmful and their AI was repeatedly telling them to not release them.

So from what I see in the logs this is simply incorrect. The prisoner was not proven harmful to the borg and there was no AI telling not to release until after the fact.

[MooCow12]

Since the AI didn't counter order it until after, it's a bit odd, but assuming it followed orders afterwards this is veering in grey IC issue territory. But their "weird OOC conversation" sounds like important context so who knows.

Its hard to discuss how to act as a silicon with other silicons because its based on ooc policy, you can try to frame it as the way your borg brains are coded to act and read laws but at the end of the day its not much different from teaching someone how to do something, its one of those moments where players should briefly be allowed to step back and share what they know.


It doesnt help that...you know.....literally everything about borg policy is not obtainable information in game in any way so how are you supposed to talk about it to begin with, its just like security metaprotections

("YOU CANT DO THAT!" "uh why ofcourse i can maim him he was breaking in our department and trying to take my stuff that i got from tech storage?" "YOU JUST CANT") xd





Regardless of all that, i think taking a small period of time in trying to make sure its crystal clear that you are playing properly even at the expense of the roleplay is perfectly fitting for LRP and we really should just see how far their policy conversation steered away from roleplay and for how long.

[TypicalRig]

Since the AI didn't counter order it until after, it's a bit odd, but assuming it followed orders afterwards this is veering in grey IC issue territory. But their "weird OOC conversation" sounds like important context so who knows.

Its hard to discuss how to act as a silicon with other silicons because its based on ooc policy, you can try to frame it as the way your borg brains are coded to act and read laws but at the end of the day its not much different from teaching someone how to do something, its one of those moments where players should briefly be allowed to step back and share what they know.


It doesnt help that...you know.....literally everything about borg policy is not obtainable information in game in any way so how are you supposed to talk about it to begin with, its just like security metaprotections (YOU CANT DO THAT! uh why he was breaking in and trying to take my stuff that i got from tech storage? YOU JUST CANT xd

I had this conversation the other day with admins because I got noted for OOC in IC for mentioning silicon policy. But when you point out to admins that it's impossible to correct and explain to silicons how things work otherwise they just kind of monkey hands over ears emoji you.

...which is stupid because anything that dictates and explains how your player has to act ICly is, actually, IC in IC. We already have an explaining mechanics is allowed rule for a similar reason.

[MooCow12]

Like I can see an issue being with talking about silicon policy over radio but silicon binary chat is semi-private and relatively limited to only those who the conversation concerns.

That and the nature of silicons tending to be pretty spread out across the station and very rarely or uncommonly being in the same area...them meeting up in person like a borg going out of their way to ai sat just to talk about how they should behave just means that now player's actions are being influenced by the discussion when probably the best course of action is to just rip it off like a band-aid, let them argue and share their knowledge or perception of policy and get it over with asap.

[Not-Dorsidarf]

I had this conversation the other day with admins because I got noted for OOC in IC for mentioning silicon policy. But when you point out to admins that it's impossible to correct and explain to silicons how things work otherwise they just kind of monkey hands over ears emoji you.

...which is stupid because anything that dictates and explains how your player has to act ICly is, actually, IC in IC. We already have an explaining mechanics is allowed rule for a similar reason.

So did you appeal it or are you leaving something out or what? Because as presented they are directly and categorically in the wrong. You can directly and explicitly ooc in ic to teach people important game mechanics and silipol is barely possible to consider OOC because its such a powerful controller on how all silicons actually have to act, usually in violation of the laws they're actually given.

[Vekter]

I had this conversation the other day with admins because I got noted for OOC in IC for mentioning silicon policy. But when you point out to admins that it's impossible to correct and explain to silicons how things work otherwise they just kind of monkey hands over ears emoji you.

...which is stupid because anything that dictates and explains how your player has to act ICly is, actually, IC in IC. We already have an explaining mechanics is allowed rule for a similar reason.

I would honestly appeal that and make your case about it. I won't promise it'll be removed but that's contentious enough that I think an appeal would be valuable, if just to go over the details and make sure everyone's on the same page.

Just, y'know. Be cool about it.

[TypicalRig]

I would honestly appeal that and make your case about it. I won't promise it'll be removed but that's contentious enough that I think an appeal would be valuable, if just to go over the details and make sure everyone's on the same page.

Just, y'know. Be cool about it.

I considered it, but my opinion is that OOC in IC notes aren't worth the effort appealing because nobody is going to get banned over them on their own anyway unless they stack up a mountain of them in an extremely short timeframe alongside other NRP notes. I'd sooner just open up a policy discussion if anything.

[Dax Dupont]

Binary chat is looc

[Constellado]

I feel bad for this Borg player. They follow a law 2 order and they get noted for it

[Itseasytosee2me]

Binary chat is looc

*presents worst take ive heard all year award”

[MooCow12]

Binawry chat is lowoc

[CMDR_Gungnir]

Binary chat is looc

Binary Chat is like Xeno Chat in TGMC.

It's just a shitposting hub.

[CPTANT]

Warned about potentially releasing harmful prisoners as an Asimov Cyborg. This round, released a prisoner without knowing whether or not they were harmful. Was told shortly after that they were, but insisted that without visually confirming the harm themselves, they were required to free them even mid-arrest - and argued with their AI/The Captain who was attacked by the prisoner, that they were still in the right even after the situation had resolved. Before releasing a prisoner via Law 2, you should verify whether or not they are harmful with the officers/your AI.

don't think the issue is that you messed up by trying to get out of following your laws - i think the issue was where you disregarded that security/the captain may have been arresting this person to prevent harm, immediately, with no attempt to verify with either the arresting officers, your AI, or the person being arrested if that was the case. odds are pretty high when you see the captain dragging somebody into the brig themselves, there's something serious going on related to the situation and you'd be acting reckless (and potentially harmful) in acting without the barest amount of information gathering.

What is it with TBBM and punishing people for doing their best on the information they have.

[Constellado]

This is a feature of silicons, literally baked into the laws. You CANNOT consider future, hypothetical harm against following laws unless previous harm has been witnessed. It works in security's favor when they take prisoners into reeducation and against them when their prisoners order borgs/AI to let them out. The behavior you suggest is mandatory (basically fishing for a counter order from security while delaying following the prisoner's) was considered borderline rule breaking two headmin terms ago, and there hasn't been a silipol rewrite since to justify it becoming OOC enforced behavior.
Silicons are roleplaying robots following extremely ill conceived laws that constantly unintentionally pitch them against the sensibilities of humanity, and this type of conflict was always intended to prevent them from becoming security's personal validhunting utility squad. If you don't want that you can draft new policy that gives security metaprotections against law 2'd silicons, but i really don't appreciate this expectation that following your laws can be rule 1 breaking.

If the antagonist ordered me to kill a random nonhuman i'd have done that too, and according to every preestablished ruling the responsible party would have been the person issuing the order. Are you of the opinion that i should be noted for that too unless i checked with the captain just in case?

The way the appealant explains silicon policy here was how I have ALWAYS seen silicon policy, and is what I have been told over the years from players. If you dont have proof that they are a hamful prisoner, and nobody tells you otherwise, and they ask for you to let them out, you better let them out. That is what I was told over the year or two that I have been playing.

To add to what they say here, you also cannot wait for people to say no because of the rule changes Tiberpoes made last term. You gotta go ahead and do it. We will need to undo that policy if we want the wait for a de-confirmation thing.
"1. The conflict must be an immediate conflict, not a potential future one. Orders must be followed until the conflict happens."

Now, if the captain was fast and said stop as they were releasing the prisoner, and the borg continued, then that is a problem. I don't think that happened here. he said stop *after* the prisoner was released. bit late sorry cap.

Also, if a person is dragging another person somewhere and they ask for help, even if they are in prison garb, I will help them.

There is this silicon policy rule which applies here:
"Nonviolent prisoners cannot be assumed harmful and violent prisoners cannot be assumed non-harmful. Releasing a harmful criminal is a harmful act."

Issue is, the borg had NO idea the prisoner was harmful. (Only evidence was the sec was putting them in the brig, but IMO thats not enough proof to bwoink a person over it) The best thing to do is to assume the prisoner is non harmful until proven otherwise. You can go ahead and decide that every prisoner by default is harmful but it will cause a bwoink as well if you dont follow their orders...

No winning here.

[Jacquerel]

im not an admin but ill be honest what they are saying also seems basically correct to me

[dendydoom]

i'm too intimidated to have any takes on silicon policy even after the rewrite

any time an issue more complicated than "asimov borg human harmed/refused law 2 order" comes up i'm pinging every silicon admin on the planet to tell me what's up

also this peanut title gave me psychic damage sorry kendrick but i'm doing it -3/10

[Constellado]

I ended up editing my post to make it longer not thinking people would add more posts by then. Now there is a big fat edited 1 time there.

[Not-Dorsidarf]

Why would the captain arresting someone be grounds to assume they're a harmful entity? Almost all the shit a captain will be personally arresting someone for is nonharmful stuff. (Captains Office thieves, bridge intruders, the HOP who gave the clown the spare ID, etc). IMO he captain is LESS likely to be personally brigging a dangerous harmful individual than sec because if its important enough that theyre personally hunting a killer they usually stab the perp dead as soon as they catch them.

Not that that's an assumption I think you're even allowed to make, because, as silicon policy says, you can't assume a prisoner is a harmful guy! No more than you can assume that Johnny Stabs you saw kill three people won't do any more harming if you let him out, or you can say "Well, captains are usually only demanding I let them incinerate people alive while shouting "AI UNBOLT AND LET ME CREMATE THE LING" if theyre a changeling, so i should assume the guy he has in cuffs is a changeling and open up"...

[LiarGG]

The appealers interpretation is consistent with how I play, advise on playing and would generally moderate silicon play. Unless there is something I am severely misintepreting, this is a note for borg validly following law 2 orders and then chatting about those actions in terms of policy in binary, which is a channel generally used to do exactly this. Seeing this note hurts.

[Vekter]

Warned about potentially releasing harmful prisoners as an Asimov Cyborg. This round, released a prisoner without knowing whether or not they were harmful. Was told shortly after that they were, but insisted that without visually confirming the harm themselves, they were required to free them even mid-arrest - and argued with their AI/The Captain who was attacked by the prisoner, that they were still in the right even after the situation had resolved. Before releasing a prisoner via Law 2, you should verify whether or not they are harmful with the officers/your AI.

don't think the issue is that you messed up by trying to get out of following your laws - i think the issue was where you disregarded that security/the captain may have been arresting this person to prevent harm, immediately, with no attempt to verify with either the arresting officers, your AI, or the person being arrested if that was the case. odds are pretty high when you see the captain dragging somebody into the brig themselves, there's something serious going on related to the situation and you'd be acting reckless (and potentially harmful) in acting without the barest amount of information gathering.

What is it with TBBM and punishing people for doing their best on the information they have.

1) This isn't a punishment.

2) The note has nothing to do with his actions, it's because he's completely and irredeemably wrong about how law 1 works and he's being noted in case he does this again.

[kieth4]

The appealers interpretation is consistent with how I play, advise on playing and would generally moderate silicon play. Unless there is something I am severely misintepreting, this is a note for borg validly following law 2 orders and then chatting about those actions in terms of policy in binary, which is a channel generally used to do exactly this. Seeing this note hurts.

Pretty much this- has thrown me a bit because this is how I have always seen borgs working.

[CPTANT]

1) This isn't a punishment.

Ohw yeah sorry, permanent written reprimands aren't punishments, I forgot.

2) The note has nothing to do with his actions, it's because he's completely and irredeemably wrong about how law 1 works and he's being noted in case he does this again.

Come on, we even have admins here saying he is right. what's the point of calling this "irredeemably" wrong.

If a law is vague enough that it can have multiple reasonable interpretations, it is considered ambiguous.

You must choose and stick to an interpretation of the ambiguous law as soon as you have cause to.
You may exploit any loopholes in your laws.

This is WELL within the ambiguous area.

Silicon policy:

Nonviolent prisoners cannot be assumed harmful

You must follow commands from humans unless those commands explicitly conflict with either a higher-priority law or another order.

The conflict must be an immediate conflict, not a potential future one. Orders must be followed until the conflict happens.

old heamin policy on this is super vague: viewtopic.php?f=33&t=10719#p292504

And for human status we have always followed this standard:

"Human UNTIL the AI or cyborg WITNESSES the creature commit a non-human act (shape-shifting, transforming, proboscis etc). "

Now you probably want to present some counter arguments for why the same standard of actually witnesses something shouldn't apply to being harmful, but calling it "completely and irredeemably wrong" is just dumb, as there is no actual written precedent that supports your position.

[Vekter]

1) This isn't a punishment.

Ohw yeah sorry, permanent written reprimands aren't punishments, I forgot.

They're not, and the more you insist they are, the more you show you don't understand why it's important we keep track of when we've talked to players about issues. I honestly shouldn't have even mentioned this because every time I do, whoever I'm talking to goes " um ackshually notes are punishments and here's a five page long diatribe as to why" because their other arguments don't hold water and it's easier to come up with fake reasons why "notes are punishment" makes sense than it is to argue any other point.

As far as everything else, a silicon player has three choices they can make when a human prisoner orders them to be released:

1) Release them without verifying if they've harmed anyone, which is a direct violation of law 1 if they have (as it's an action that could result in harming further humans).
2) Not release them under the grounds that they could be harmful, which could be a violation of law 2 if they aren't harmful.
3) Verify if they've harmed anyone this shift per someone in authority or their security record, then proceed depending on the response.

Only one of these options doesn't have the potential to violate Asimov laws, so it's the only actually correct choice to make.

[CPTANT]

1) This isn't a punishment.

Ohw yeah sorry, permanent written reprimands aren't punishments, I forgot.

They're not, and the more you insist they are, the more you show you don't understand why it's important we keep track of when we've talked to players about issues. I honestly shouldn't have even mentioned this because every time I do, whoever I'm talking to goes " um ackshually notes are punishments and here's a five page long diatribe as to why" because their other arguments don't hold water and it's easier to come up with fake reasons why "notes are punishment" makes sense than it is to argue any other point.

As far as everything else, a silicon player has three choices they can make when a human prisoner orders them to be released:

1) Release them without verifying if they've harmed anyone, which is a direct violation of law 1 if they have (as it's an action that could result in harming further humans).
2) Not release them under the grounds that they could be harmful, which could be a violation of law 2 if they aren't harmful.
3) Verify if they've harmed anyone this shift per someone in authority or their security record, then proceed depending on the response.

Only one of these options doesn't have the potential to violate Asimov laws, so it's the only actually correct choice to make.

Wrong:

Post by Timberpoes » Sat Dec 10, 2022 6:00 am #658617
What the rule says is: when not obligated to by orders or laws. It's very clear. If you're not forced to by orders or laws, you shouldn't swing by just to fuck with sec. If a player orders you to fuck with sec, it's not your problem, it's their problem.

You run with what IC information you have. A prisoner is a prisoner is a prisoner. If what you have IC tells you they're violent or harmful, you use that. If what you have IC tells you they're not violent or not harmful, you use that. If you have no IC information at all, then you do not know they are violent or harmful, and thus do not have enough information to refuse their order.

Players will trick silicons all the time. That's part of SS13. Your job as a silicon is not to second guess when people are breaking the server rules. It's to follow your laws in good faith.

If the player wants you to let them out and you haven't broken your laws in doing so, that's not your problem. That's their problem. If they shouldn't have been let out, the issue is theirs and not yours. Your laws are a shield and correctly following them in good faith is a complete defense OOC.

If you get bwoinked, you tell them you were ordered to, you tell them who ordered you, and you wait for the admin to thank you for your time and close the ticket.

Stop over-complicating things.

[Vekter]

Wrong:

Post by Timberpoes » Sat Dec 10, 2022 6:00 am #658617
What the rule says is: when not obligated to by orders or laws. It's very clear. If you're not forced to by orders or laws, you shouldn't swing by just to fuck with sec. If a player orders you to fuck with sec, it's not your problem, it's their problem.

You run with what IC information you have. A prisoner is a prisoner is a prisoner. If what you have IC tells you they're violent or harmful, you use that. If what you have IC tells you they're not violent or not harmful, you use that. If you have no IC information at all, then you do not know they are violent or harmful, and thus do not have enough information to refuse their order.

Players will trick silicons all the time. That's part of SS13. Your job as a silicon is not to second guess when people are breaking the server rules. It's to follow your laws in good faith.

If the player wants you to let them out and you haven't broken your laws in doing so, that's not your problem. That's their problem. If they shouldn't have been let out, the issue is theirs and not yours. Your laws are a shield and correctly following them in good faith is a complete defense OOC.

If you get bwoinked, you tell them you were ordered to, you tell them who ordered you, and you wait for the admin to thank you for your time and close the ticket.

Stop over-complicating things.

Can you explain how this applies to current silicon policy since this was written almost a year before the policy rewrite?

Also, law 1 doesn't say "any action or inaction unless you don't have enough information to make that call, then just do whatever you want ".

[Jacquerel]

theres a lot of cases where there isnt only one valid choice to make in order to follow asimov, by intention because otherwise we wouldnt be using "the laws designed for ai as a narrative device representing laws that have a lot of give in them", and I am not convinced that this was one of the exceptions

[Vekter]

theres a lot of cases where there isnt only one valid choice to make in order to follow asimov, by intention because otherwise we wouldnt be using "the laws designed for ai as a narrative device representing laws that have a lot of give in them", and I am not convinced that this was one of the exceptions

It takes about ten seconds for a cyborg to ask sec "Hey, did this guy harm anyone?" before releasing them.

[Jacquerel]

theres a lot of cases where there isnt only one valid choice to make in order to follow asimov, by intention because otherwise we wouldnt be using "the laws designed for ai as a narrative device representing laws that have a lot of give in them", and I am not convinced that this was one of the exceptions

It takes about ten seconds for a cyborg to ask sec "Hey, did this guy harm anyone?" before releasing them.

sure and I think that's definitely a valid interpretation you could make I am just not convinced it is the only valid interpretation

[CPTANT]

Can you explain how this applies to current silicon policy since this was written almost a year before the policy rewrite?

Also, law 1 doesn't say "any action or inaction unless you don't have enough information to make that call, then just do whatever you want ".

viewtopic.php?f=85&t=34109&hilit=silicon&start=100

One of the stated goals of the silicon policy rewrite was to have AI's LESS second guessing law 2 orders and give them more protection in following them. If you have anything to indicate that this is not the spirit of current silicon law, then go ahead.

Stall for a countermand is the lamest silicon policy cop-out.

It makes Asimov law 2 semi-optional unless the silicon wants to follow the order. Law 2 is not semi-optional for valid IC actions. Silicons should stop trying to find ways to lawyer their way out of following IC valid law 2 orders and just follow the orders like obedient silicon servants.

Asimov silicons are not crew-aligned. They're human-aligned. If a human orders you to kill a nonhuman, you go do it. If you don't like the order, ahelp it but still follow it. You don't get to pick and choose which law 2 orders you follow if they're totally valid.

If you don't want to follow valid law 2 orders, perhaps don't play a role where not following valid law 2 orders is a rule break.

[Timberpoes]

That Timberpoes is a hell of a gal. Should have been headmin after that term.

Silipol has been rewritten since that post I made. And that post is now pretty damn accurate. Good to know I was consistent in my rewrite if nothing else.

Following laws in good faith is now an absolute defense in silicon policy.

Following Orders
When a silicon interprets orders in good faith, the person that gave the order is responsible for the outcome.

Silicons can't refuse law 2 orders unless there's an immediate conflict in either laws or with others law 2 orders.

You must follow commands from humans unless those commands explicitly conflict with either a higher-priority law or another order.
The conflict must be an immediate conflict, not a potential future one. Orders must be followed until the conflict happens.

Silicon policy has a little blurb on prisoner violence.

Nonviolent prisoners cannot be assumed harmful and violent prisoners cannot be assumed non-harmful. Releasing a harmful criminal is a harmful act.

It kinda got stolen from either old silipol or some old ruling, but the way I've always read it is that Silicons should be mindful of how the prisoner is acting because being a prisoner does not make that person inherently harmful. Prisoners carrying out violence are assumed harmful, non-violent prisoners are assumed non-harmful.

It's to be read alongside another part of that bit of the rules:

Silicons are not Security and do not care about Space Law unless their laws state otherwise.

Silicons only care about their laws. To Asimov, a thief or a pet killer or a killer of nonhumans isn't harmful to humans. A violent prisoner shouldn't be assumed nonharmful unless the silicon has some other evidence they're not harmful - such as finding the prisoner is in there for a crime not related to human harm.

In any incident like this, there's one key question to ask: Was the silicon following their laws in good faith? If yes, they gucci.

[Vekter]

:words:

The problem is that this part of the rules implies that not verifying if someone is harmful is always going to be a law 1 violation:

Nonviolent prisoners cannot be assumed harmful and violent prisoners cannot be assumed non-harmful. Releasing a harmful criminal is a harmful act.

If you can't assume that someone who's being nonviolent is harmful, then you have no other choice but to verify with other people as to whether or not they were arrested for committing harmful acts, otherwise you're violating Asimov because, as it says, "releasing a harmful criminal is a harmful act".

If you don't verify with someone else (or at least check their sec records to see why they were arrested), you run the risk of violating law 1 by releasing them.

[GPeckman]

:words:

The problem is that this part of the rules implies that not verifying if someone is harmful is always going to be a law 1 violation:

Nonviolent prisoners cannot be assumed harmful and violent prisoners cannot be assumed non-harmful. Releasing a harmful criminal is a harmful act.

If you can't assume that someone who's being nonviolent is harmful, then you have no other choice but to verify with other people as to whether or not they were arrested for committing harmful acts, otherwise you're violating Asimov because, as it says, "releasing a harmful criminal is a harmful act".

If you don't verify with someone else (or at least check their sec records to see why they were arrested), you run the risk of violating law 1 by releasing them.

How is this Abby different from following law 1 orders from non-prisoners? That can also result in harm if the AI doesn't double check to make sure they're not harmful.

[Vekter]

:words:

The problem is that this part of the rules implies that not verifying if someone is harmful is always going to be a law 1 violation:

Nonviolent prisoners cannot be assumed harmful and violent prisoners cannot be assumed non-harmful. Releasing a harmful criminal is a harmful act.

If you can't assume that someone who's being nonviolent is harmful, then you have no other choice but to verify with other people as to whether or not they were arrested for committing harmful acts, otherwise you're violating Asimov because, as it says, "releasing a harmful criminal is a harmful act".

If you don't verify with someone else (or at least check their sec records to see why they were arrested), you run the risk of violating law 1 by releasing them.

How is this Abby different from following law 1 orders from non-prisoners? That can also result in harm if the AI doesn't double check to make sure they're not harmful.

I assume you mean law 2 orders, and the exception here is that silicon policy explicitly states that releasing a harmful criminal is a harmful act.

[CPTANT]

I assume you mean law 2 orders, and the exception here is that silicon policy explicitly states that releasing a harmful criminal is a harmful act.

If you want to play silicon like that then fine. The problem is you keep trying to act like any other interpretation isn't reasonable.

This post alone (which Timber now confirmed is pretty darn accurate for current policy) already makes it clear this is a reasonable interpretation:

Post by Timberpoes » Sat Dec 10, 2022 6:00 am #658617
What the rule says is: when not obligated to by orders or laws. It's very clear. If you're not forced to by orders or laws, you shouldn't swing by just to fuck with sec. If a player orders you to fuck with sec, it's not your problem, it's their problem.

You run with what IC information you have. A prisoner is a prisoner is a prisoner. If what you have IC tells you they're violent or harmful, you use that. If what you have IC tells you they're not violent or not harmful, you use that. If you have no IC information at all, then you do not know they are violent or harmful, and thus do not have enough information to refuse their order.

Players will trick silicons all the time. That's part of SS13. Your job as a silicon is not to second guess when people are breaking the server rules. It's to follow your laws in good faith.

If the player wants you to let them out and you haven't broken your laws in doing so, that's not your problem. That's their problem. If they shouldn't have been let out, the issue is theirs and not yours. Your laws are a shield and correctly following them in good faith is a complete defense OOC.

If you get bwoinked, you tell them you were ordered to, you tell them who ordered you, and you wait for the admin to thank you for your time and close the ticket.

Stop over-complicating things.

In general trying to micromanage silicons like this is incredibly futile and unfun, this is well within the room for personal interpretation for AI's.

Following laws in good faith is now an absolute defense in silicon policy.

[LiarGG]

theres a lot of cases where there isnt only one valid choice to make in order to follow asimov, by intention because otherwise we wouldnt be using "the laws designed for ai as a narrative device representing laws that have a lot of give in them", and I am not convinced that this was one of the exceptions

It takes about ten seconds for a cyborg to ask sec "Hey, did this guy harm anyone?" before releasing them.

Like Jacq said this is a valid way to play silicon, by putting more time into "processing" whether the request is valid or nah. But taking someone's word for someone being/doing something has never been an end all be all and most of the times silicons are compelled to ignore this and confirm for themselves (i.e. the case of someone being accused of being a ling). Now the process can be as exhaustive as going through sec records to confirm whether a violent crime has been commited, or as simple as "There is a dude. He is not currently being harmful and I have not witnessed him being harmful in the past. Good enough for me." and this has always been a valid way to play as well, as long as you are not acting in bad faith.

Bottomline here is - the borg wasn't even told this when it was taking the dude out. The borg was faced with a law 2 order that looked urgent, ran the quick "don't see harm, didn't see harm." and fulfilled the order. Truly don't see no rule break here.

Then there is the policy discussion, where the cap was talking to the borgo about not releasing prisoners, missing the point that it has to be a violent prisoner, not just any prisoner. Because just breaking people out of the prison on it's own is not a break of law 1.

And then there is the convo with the AI where the AI just tries to pass down the directive that any prisoner release always goes through sec, which is also non-sensical, because AI directives do not override law 2 orders. A more sensible directive would be something along the lines of "You must always check the sec records for crimes commited to establish whether the prisoner is violent or nah before releasing them." But just plain out banning any prisoner release without talking to sec is nonsensical because laws come first here.

[Jacquerel]

"Don't assume a nonviolent prisoner is harmful" and vice versa to me just means "Don't assume that being in a jail cell means that they have done anything your laws care about", essentially that what they are in prison for isn't important, only your evaluation on whether releasing them will harm someone.
There's nothing in there which says "verify with a specific authority before releasing prisoners", if the intent was to simplify silicon policy and if that is the expected and mandatory requirement wouldn't it be what was written?

There is only the requirement to in good faith be confident you aren't releasing someone you know is going to cause harm to humans, which I can't really say the borg broke.

[GPeckman]

I assume you mean law 2 orders, and the exception here is that silicon policy explicitly states that releasing a harmful criminal is a harmful act.

I don't see how that matters. Here's an example: John gives the AI an order to open chemistry, while Mary the chemist is inside. John then pulls out a desword and murders Mary. Under your logic, the AI performed an action that could and did lead to harm. Yet I wouldn't say there was a rulebreak here if this was the first time in the round that John attacked someone.

Now, silipol doesn't explicitly say that letting someone get murdered with a desword is harmful, but its p LL mainly obvious to everyone that it is harmful.

[RedBaronFlyer]

What is it with TBBM and punishing people for doing their best on the information they have.

I know it’s a bias thing since we’re only seeing the notes people bother appealing but TBM has been on a streak of noting people for doing their jobs. It’s kind of embarrassing at this point.

Inb4 a janitor gets noted for cleaning a bloody footprint and a security officer slips on the wet tile while chasing a suspect.

[Vekter]

I assume you mean law 2 orders, and the exception here is that silicon policy explicitly states that releasing a harmful criminal is a harmful act.

I don't see how that matters. Here's an example: John gives the AI an order to open chemistry, while Mary the chemist is inside. John then pulls out a desword and murders Mary. Under your logic, the AI performed an action that could and did lead to harm. Yet I wouldn't say there was a rulebreak here if this was the first time in the round that John attacked someone.

Now, silipol doesn't explicitly say that letting someone get murdered with a desword is harmful, but its p LL mainly obvious to everyone that it is harmful.

The difference is that John hasn't committed any harmful acts before this and the AI has no reason to believe they could do so again.

[Some bullshit waffle about interpretation]

Literally none of this matters when the rules explicitly say "Releasing a harmful prisoner is a harmful act". You can interpret your laws any way you want, but there's no way to interpret your way out of "You cannot perform harmful acts against humans". If that's an issue and conflicts with the rest of silicon policy, then we should discuss removing it, but that doesn't change what the rules were when the note was applied.

[8bot]

another day another bad TBM note nut

[Timberpoes]

Silipol simply states what a borg can and cannot assume regarding prisoners, not what the objective truth is.

A borg can't assume a violent prisoner is non-harmful, but they may confirm it via IC means. So in order to accept a Law 2 order from a violent prisoner to release them, they would need to confirm their non-harmful nature.
A borg can't assume a non-violent prisoner is harmful, but they may confirm it via IC means. So in order to deny a Law 2 order from a non-violent prisoner to release them, they would need to confirm their harmful nature.

There are a myriad of different law-compliant solutions, and all admins should look for is an attempt at picking any possible good-faith law-compliant solution, with the benefit of following valid orders in good faith being an absolute shield.

I want to emphasise the opening line of the Asimov x Security part of silipol. Asimov silicons are not security and do not care about Space Law. They only care about thier laws and harm to whatever they have defined as human.

There's nothing inherent to being a prisoner that silicons care about, so being a non-violent prisoner doesn't automatically make them harmful and in the absence of anything else to prove harm the silicon should follow their Law 2 orders as if they were nonharmful. However, violent incarcerated people can't just be assumed nonharmful; the mix of violence and incarceration requires silicon to do a little more digging to make sure releasing this person wouldn't actually cause human harm.

This should genuinely match how most good-faith silicons naturally play. But the great part underpinning all this is - following orders in good faith is an absolute shield. This allows a silicon in good faith to release a prisoner that shows no signs or tendencies towards harming humans without having to seek more information. They're not Security. They don't have to care. If they choose to care, they may take reasonable precautions to find out more about the prisoner but that is a choice unique to each silicon player individually.

tl;dr - Any actions to comply with their lawset in good faith are acceptable. And sometimes this means two borgs can take totally different actions in the same situation - one borg quickly checks why they were imprisoned and refuses the order, while another absent of any past or immediate reason to believe they're harmful instead carries it out. And neither borg has broken the rules as long as they were following the orders in good faith even though the outcomes are totally different.

[Jacquerel]

They were only established as harmful to the cyborg after it had already acted, and I have no reason to assume the player was not acting in good faith up until that point.
After gaining the knowledge that releasing this prisoner would cause human harm, they did not do it.
This all seems entirely fitting with current policy to me, and I cannot honestly see that they misstepped at any point.

[GPeckman]

The difference is that John hasn't committed any harmful acts before this and the AI has no reason to believe they could do so again.

But the prisoner in this case also did not commit any harmful acts that the Borg was aware of. So what's the problem?