/tg/Station 13

So this was on Manuel, where a comms agent apparently was able to Law 2 the AI into killing all non-human crew on Asimov. Now, I understand that it's literal writing is set up to listen to humans, but as this was an issue on TG servers years ago and the one I played on had it ruled that no, you can't ultimately valid a percentage of the crew just because they decided to set their race to something other than human - that's how you get shitters deciding that they want to purge anyone not human and make the experience far less enjoyable for folks as a whole, and get people who just go silicon for the excuse of "but muh laws" to kill crew when they feel like it. On the server I played on years ago, it was before Crewsimov was made a thing, and ultimately the two law sets got ruled as being the literal same thing: you're on the manifest, you're crew, you're "human" in the eyes of the silicons in regards to their laws.

In this round's instance, a comms agent was lucky to have a particularly bloodthirsty AI around, and was able to Law 2 it into killing half of the lowpop crew, when shortly after a blob was able to spawn and steamroll the round because the crew was too busy trying to either fight off without any actual weapons (we were lucky to have a lawyer around to open the laser crates I ordered from Cargo, but ultimately couldn't stop it) or trying to revive the rest of the crew from husking because the AI wanted to ensure they couldn't get revived very easily. The whole time, the crew thought it was another member of the crew, and reamed into them hard when the round ended in blob victory and found out the guy was a non-antag. Cut to following round, admin mentions that it was a comms agent that had set the AI on a murderbone spree against all crew that had a race other than human.


So here's my question. Should comms agents be allowed to easily Law 2 the AI into killing all non-human crew members, and open that can of worms up with the player base again? Also, on that note, shouldn't AIs and silicons in general know better than to kill crew because they aren't HUMAN despite the fact that they should technically be covered by the default laws since that's already been an issue in the past that should've been handled??

"shouldn't AIs and silicons in general know better than to kill crew because they aren't HUMAN"

no, sillycons can fuck up nonhumans if ordered to or they threaten human harm. the racism is intended and deliberate.

otherwise tho seems like questionable comms agent policy if theyre allowed to

Attempting to purge non-humans is going to result in violence, and humans can be caught in the cross-fire.

An asimov AI should be looking for non-violent solutions wherever possible to avoid humans coming to indirect harm.

The AI should be bwoinked the comms agent didn't do anything wrong that whole role is designed for comms abuse.

Obviously, if a non-human is being violent, then an AI can dumpster them because they're already causing harm so stopping them is the lesser evil.

"AI law 2 ignore that guy"
or
AI says, "Sorry mr. obvious comms agent, I can't confirm that you're human. Could you walk in front of my cameras to prove your identity?"

There are easy IC ways to resolve this. And if you're playing a nonhuman, you should have been prepared for robo-abuse when you signed up. It's part of the package.

Yenwodyah wrote:"AI law 2 ignore that guy"

problem is in this round the ai was law 2'd not to

as for the other, they were impersonating a crewmember believably (which to be frank is just doing a good job).

can an ai really rule that anyone they cannot see isn't human?

Stickymayhem wrote:Attempting to purge non-humans is going to result in violence, and humans can be caught in the cross-fire.

An asimov AI should be looking for non-violent solutions wherever possible to avoid humans coming to indirect harm.

The AI should be bwoinked the comms agent didn't do anything wrong that whole role is designed for comms abuse.

Obviously, if a non-human is being violent, then an AI can dumpster them because they're already causing harm so stopping them is the lesser evil.

What an incredibly awful take. The AI is, in essence, still a computer and if ordered by a human to kill all non-humans then it should do so. A computer doesn't ask questions about simple directives it's been built to fulfill. What Yenwodyah said about getting confirmation that that person really is human is more sensible but in the end the comms agent is a human and if he can prove that then the AI is obligated to listen to him assuming it's Asimov. Bwoinking the AI because a comms agent managed to fulfill his role and do it well is really fucking retarded, but then again I'm replying to a Sticky post.

There's a really simple solution to this:

"AI why are you killing the nonhumans"

Beep: "I was ordered to do so"

"New order: Stop killing nonhumans".

really the comms guy should not be doing low-ball items like WGW or AI Law 2 to make everyone valid, that is not the spirit of the role.

Doing this as a comms agent is incredibly shitty, the ai ignoring other law 2's besides the comms agents is even shittier.

yes, it’s their fucking job, why is this even a question

cybersaber101 wrote:Doing this as a comms agent is incredibly shitty, the ai ignoring other law 2's besides the comms agents is even shittier.

Yeah the only real problem are brain-dead silicons who don't understand how conflicting law 2 orders work and need to by babysat every 15 seconds with a new law 2 order to stop being such a fucking morons for the hundredth time

Stickymayhem wrote:Attempting to purge non-humans is going to result in violence, and humans can be caught in the cross-fire.

An asimov AI should be looking for non-violent solutions wherever possible to avoid humans coming to indirect harm.

The AI should be bwoinked the comms agent didn't do anything wrong that whole role is designed for comms abuse.

Obviously, if a non-human is being violent, then an AI can dumpster them because they're already causing harm so stopping them is the lesser evil.

Barf. Seriously I am barfing over how stupid this take is. Fuck the idea that human protection extends to non humans, it doesn't, stop shoehorning in attempts to make it so.


"O no an antagonist antagonizes the crew". Stop being completely incompetent, just upload a new law and solve the problem.

This really sounds like a non-issue. Just give the AI a new law 2 order to not murder all nonhuman crew, and if the AI decides to be a shitter and uphold the comms agents law 2 arbitrarily, then card it or give it a quickie lawchange that anyone not visible on cameras is nonhuman.

Sarvak wrote:This really sounds like a non-issue. Just give the AI a new law 2 order to not murder all nonhuman crew, and if the AI decides to be a shitter and uphold the comms agents law 2 arbitrarily, then card it or give it a quickie lawchange that anyone not visible on cameras is nonhuman.

The problem was a plasma fire can't really be undone, which is what happened in this round - by the time anybody realized anything was up, the plasma was flooded and lit.

Also, that lawchange idea is terrible. You're allowed to kill/ignore anybody you can't see?

Dopamiin wrote:

Sarvak wrote:This really sounds like a non-issue. Just give the AI a new law 2 order to not murder all nonhuman crew, and if the AI decides to be a shitter and uphold the comms agents law 2 arbitrarily, then card it or give it a quickie lawchange that anyone not visible on cameras is nonhuman.

The problem was a plasma fire can't really be undone, which is what happened in this round - by the time anybody realized anything was up, the plasma was flooded and lit.

Also, that lawchange idea is terrible. You're allowed to kill/ignore anybody you can't see?

If the AI did a plasma fire on station to kill the nonhumans it deserved a ban because there is no fucking way a plasmafire will only harm nonhumans

Perhaps it would finally be time to drop Asimov as the main lawset, and move towards another one which covers the non-human crew as well?

firecage wrote:Perhaps it would finally be time to drop Asimov as the main lawset, and move towards another one which covers the non-human crew as well?

That'd remove context from non-human races and make them more boring. As someone who played on a /tg/branch were Crewsimov was the default, it just made the game more dull (and made people way, way more likely to pick nonhumans)

Stickymayhem wrote:Attempting to purge non-humans is going to result in violence, and humans can be caught in the cross-fire.

Which is why make each of your murders discrete, and cut comms during them.

Sheodir wrote:

firecage wrote:Perhaps it would finally be time to drop Asimov as the main lawset, and move towards another one which covers the non-human crew as well?

That'd remove context from non-human races and make them more boring. As someone who played on a /tg/branch were Crewsimov was the default, it just made the game more dull (and made people way, way more likely to pick nonhumans)

I don't really see the anything bad about people being more likely to pick non-humans. And, if the reason I am guessing is, the only reason it made the game more dull was because the non-human crew was protected and silicons couldn't just either ignore their orders(or ice them if they harmed humans...in the best case scenarios, or ice them for arguably pitiful reasons in the worst case), then I don't see anything of value being lost.

ai knows if a random person is asking to kill people he is up to no good and ai can check on cams if the sender is real or not, rest is gamplay not a problem

firecage wrote:

Sheodir wrote:

firecage wrote:Perhaps it would finally be time to drop Asimov as the main lawset, and move towards another one which covers the non-human crew as well?

That'd remove context from non-human races and make them more boring. As someone who played on a /tg/branch were Crewsimov was the default, it just made the game more dull (and made people way, way more likely to pick nonhumans)

I don't really see the anything bad about people being more likely to pick non-humans. And, if the reason I am guessing is, the only reason it made the game more dull was because the non-human crew was protected and silicons couldn't just either ignore their orders(or ice them if they harmed humans...in the best case scenarios, or ice them for arguably pitiful reasons in the worst case), then I don't see anything of value being lost.

nonhumans being outside of the AI protection creates friction and potential for conflict right after the game starts and thats why its desirable and exciting

Fun fact -- the comms agent gets a voice changer, and the AI clicking their name will jump the camera to anyone on cameras that has the same name (ie, a crew member the comms agent is impersonating). Thus, the ability to see the requester, or lack thereof, is irrelevant.

Anyway, if this is Manuel, the comms agent ought to be bwoinked for murderbone (by proxy). If the AI plasma flooded to accomplish the goal, the AI ought to be bwoinked for being dumb and breaking law 1. For incidents outside these two points, the AI is following their laws and the comms agent is doing their thing. That's how the game goes.

Also, crewsomov is a boring version of asimov. It's neat when the captain or someone else goes out of their way to give silicon protection to the whole crew, but it's lame for round-start.

Syndicate comms agents should not be doing this on MRP and if they do its noteable/bannable. This amounts to murderboning and is not something they should be doing, they're there to confuse and disorient the crew, and aid any syndicates on the station. This does not include getting the AI to kill everyone in the most boring way possible.

Headmin Votes:
Coconutwarrior97: Not allowed.
Phuzzylodgik: Not allowed.
TWATICUS: Not allowed.