Print

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

[ThtblovesDF]

The soap in the showers is what will get you out. Place it on the door, stack random stuff on it and riot wildly, once someone enters, choke them out, take there id, cuff them to a bed, put on there pants and walk out fast enough that none can glance at you for to long. By the time the next person checks on the perma prison (or the guard undoes his cuffs), you will hopefully be gone.

Alternativly, you can destroy electric grills by using a glass shard on them.

I personally quite enjoy preparing a exit way (turn a wall into a slide wall with some tools behind it), but doing that is a challange on its own... then again, none expects you to break into a empty prison.

[scrdest]

Changing subject.  Any thoughts on AI rulings?  Two things in particular:

1: Last night, a non-Malf AI bolted xeno and refused to open it.  We ended up carding him.  Because slimes can be dangerous, he absolutely refused to let anyone in, even the research staff.

2: Last round, I played AI and the CE built walls around doors because atmosphere was leaking.  The doors were already bolted shut and the captain had gone into space (with hardsuit) to find someone.  The CE walked up and walled it over so the captain couldn't get back in except through maintenance.  I suggested giving the CE a few baton hits and people start saying I'm Malf for suggesting harm - there's a difference between harm and discipline.

Clearly we need some house rules on exactly how much that First Law governs anything.  I mean, if we're gonna take it to the extreme then during an AI round I should just bolt every door.  Humans cannot starve, but they may injure themselves or each other, therefor locking them in small closets is the most ideal situation for protection.

People take the laws far too literally sometimes...

I'd say 1 is being deliberately annoying; in some discussion about that law in SS13 (maybe even here), someone argued that 'harm' is not what is potentially harmful - which, to be fulfilled, would require you to bolt not only Toxins and Xenobio, but also Bar and Kitchen - alcohol is harmful, eating to much is harmful - Chemistry - all the drugs which can be made - Engineering - SINGULO IS LOOSE ABANDON SHIFT!---

And so on. If you'd accept the 'all potential future harm' option, a self-consistent AI should quickly disappear in a puff of logic due to the protection itself being potentially harmful.

On the other hand, harm being 'present harm' means no, Xenobio is not harmful RIGHT NOW, nobody is or about to be eaten, do not bolt. An assistant with a katana slashing wildly is present potential harm, i.e. someone is not being actually harmed this second, but only because the blade slipped a bit, AI should do what it can to non-lethally stop the assistant. A criminal escaping the brig is a bit hard to decide, but most AIs would probably detain him.

2. No matter if you punch someone because you want to discipline them or because you're a sadist, it IS harm to the person being hit. Also, keep in mind that we're very laissez-faire about applying electricity to people, which IRL is a very very bad idea.

[Urist_McGamer]

I mean, if we're gonna take it to the extreme then during an AI round I should just bolt every door.  Humans cannot starve, but they may injure themselves or each other, therefor locking them in small closets is the most ideal situation for protection.

Personally I think that the state of the station and staff at the start of the round should be presumed safe by the AI.  The AI should then respond reactively to the events that unfold, and changes to the station.  Proactively locking down the station, while it is a possible interpretation of the Laws, falls into the "not fun for anyone else" category.   All jobs can be done safely if done correctly.  The AI should presume that the people in that dept are capable of performing the job correctly until proven otherwise.

Due to the literal interpretation of laws by so many people, I'd suggest to push in a law somewhere, probably right between 1 and 2, "Any crew certified to a job is considered safe in such environment until emergency requires aid."

I seem to recall a story by Asimov dealing with something like that. Robots working alongside humans in some mine environment kept destroying themselves trying to keep human workers from exposing themselves to low levels of radiation. I don't remember exactly, but the solution might have been to modify rule one.

[Hanslanda]

I escaped from permabrig with a fork. Granted, a random event helped me out here, but still. A fork.

Gr3y T1de virus?

Nope. I'm not going to tell you either. It happened though.

Also, play the arcade machine. Get the singularity toy. Put the soap down, put the toy on top of it. Get the guard to come in, over the toy. He'll slip, you jack his baton/what have you, steal his stuff, then escape. Hell, the Arcade machine can give you a full set of alternate clothing to help disguise your appearance.

[Girlinhat]

2. No matter if you punch someone because you want to discipline them or because you're a sadist, it IS harm to the person being hit. Also, keep in mind that we're very laissez-faire about applying electricity to people, which IRL is a very very bad idea.

I suppose the real question is, "How much harm is harmful?"  AI allows public floggings, but avoids slime-related-face-melting.  There's levels of harm here, from discipline to fatal.  When I play AI, I class 'harm' as 'something harmful enough to be lethal, crippling, or leave the individual vulnerable to other things'.  As such, a stun baton is fine.  A harm baton is fine.  A harm baton used by the clown is not fine, because the clown is likely going for blood, and the victim is vulnerable to others taking advantage of the situation.  A security officer can use generally any weapon against humans because they know how to handle it and won't let the victim come to more harm than required.

When you get into law enforcement, the AI would end up shutting down the brig entirely.  Unless you clarify 'how much harm is harmful' and the AI allows injury.

[scrdest]

2. No matter if you punch someone because you want to discipline them or because you're a sadist, it IS harm to the person being hit. Also, keep in mind that we're very laissez-faire about applying electricity to people, which IRL is a very very bad idea.

I suppose the real question is, "How much harm is harmful?"  AI allows public floggings, but avoids slime-related-face-melting.  There's levels of harm here, from discipline to fatal.  When I play AI, I class 'harm' as 'something harmful enough to be lethal, crippling, or leave the individual vulnerable to other things'.  As such, a stun baton is fine.  A harm baton is fine.  A harm baton used by the clown is not fine, because the clown is likely going for blood, and the victim is vulnerable to others taking advantage of the situation.  A security officer can use generally any weapon against humans because they know how to handle it and won't let the victim come to more harm than required.

When you get into law enforcement, the AI would end up shutting down the brig entirely.  Unless you clarify 'how much harm is harmful' and the AI allows injury.

How is Harmbaton fine? Its sole purpose is inflicting physical harm. If you need to stop someone, a stun will do. What is the purpose of hurting people as non-traitor?

[Glloyd]

meanwhile Permabrig is KTHANKSBYEROUNDLOCKOUT.

Not really, unless you're straightjacketed. The new permabrig is DESIGNED for escape attempts, and since its introduction, has seen MULTIPLE escape attempts. Some sucessful, others not. Some insane, some not. Well, mostly insane.

Seriously? I didn't get to see any, and cannot think of anything aside from being broken out.

You cannot try the northern windows, electrified. If someone opens the door, there is still another door behind it, and then Brig, which is generally not the place you want to escape to. If you want to try to get to Armory, you need a whole lot of tools and time to breach the walls. At least half of the other walls cover void. Ideas?

meanwhile Permabrig is KTHANKSBYEROUNDLOCKOUT.

Not really, unless you're straightjacketed. The new permabrig is DESIGNED for escape attempts, and since its introduction, has seen MULTIPLE escape attempts. Some sucessful, others not. Some insane, some not. Well, mostly insane.

I escaped from permabrig with a fork. Granted, a random event helped me out here, but still. A fork.

Gr3y T1de virus?

Break the grill with a piece of glass from the broken window, which you break using one of the emergency oxygen tanks. Superspacewalk to the external construction site and live, albiet in worse shape.

Alternate: In the top cell, there's a shank. Shank your way out.

Alternate: (This has happened in a rev round) Gather your inmates and bumrush the door when someone opens it. Orange tide anyone?

Alternate: USE YOUR IMAGINATION

Alternate: Hope for a random event (I've seen people escape due to wormholes, immovable rods and meteors, to name a few)

Alternate: Break a window, complain about air loss to the AI, when he sends an engineer in, SHANK SHANK SHANK. Or disarm. Grab his tools and hack your way out.

Alternate: Be cool and play the video games in there. Win some fun stuff.

[andrea]

I seem to recall a story by Asimov dealing with something like that. Robots working alongside humans in some mine environment kept destroying themselves trying to keep human workers from exposing themselves to low levels of radiation. I don't remember exactly, but the solution might have been to modify rule one.

read that recently. Yes, they weakened the first law, by leaving out the part about "or through inaction etc.", which however resulted in the robots being unstable.
Which, honestly, would explain a lot about SS13 AIs

[kg333]

The entire point of Asimov's laws is that they are full of holes...the man only wrote several books about all the problems with them.

I will add interpretation of the laws should fall under the "keep things fun for everyone" principle, granted a fully functional AI.  For example, locking everyone out of Xeno because "slimes might cause harm to qualified personnel" isn't really fun for anyone but AI, and shouldn't be happening unless AI were given some appropriately ridiculous instruction by the heads ("AI, assume all non-human life extremely dangerous!" in a changeling round).  Should AI be Malf...well, traitors have latitude.

KG

[scrdest]

The entire point of Asimov's laws is that they are full of holes...the man only wrote several books about all the problems with them.

I'm totally stealing that idea. I'll come up with a faulty AI lawset and then get rich writing elaborate stories explaining how did I cock up the lawset. [Insert evil laughter]

[BigD145]

Alternate: Hope for a random event (I've seen people escape due to wormholes, immovable rods and meteors, to name a few)

If you're the only antag left and things are settling down I would hope the admins would random event and let the RNG kill/assist permabrig.

Alternate: Break a window, complain about air loss to the AI, when he sends an engineer in, SHANK SHANK SHANK. Or disarm. Grab his tools and hack your way out.

What did that engineer ever do to you? They tried to fix your loss of air, that's what. Don't shank the engineers.

[10ebbor10]

Anyway, a question I keep forgetting the answer too.

Are the AI's law tiered. Ie, does law 1 overrule law 2,...
Many players rely on them to be, but only Asimov and Antimov have it exlicitedly stated.

Also, if they are not tiered, am I right to assume they are tiered in initialization. Ie, law 1 loads before law 2, so if you got law 1 which says it overrides all laws, and law 2 which tells the AI to ignore law 1, law 1 wins.

[TheBronzePickle]

AI laws are not tiered. In fact, the corporate lawset depends entirely on the laws being all equal in value. So, unless otherwise specified, treat all laws the same.

[Jacob/Lee]

TG wiki says laws are tiered, yes. At least Asimov. This is what makes the hacked AI board not useless, since the law will be placed before every other and you can make it anything you please.

[Karlito]

One way to interpret the wiki gives Hacked modules precedence over Core Laws and Core Laws precedence over Freeform Laws, without Core Laws necessarily having precedence amongst themselves, though that might be going out on a bit of a limb. I think the consensus from the last discussion was to let the AI player work out these details themselves, anyway.