|
Why is the AI so special and why are AI players such babies? I approach (and think everyone else should too) playing a round on goonstation with a 80% certainty that i'm going to die and not see the round out. The AI has an all-seeing x-ray eye, so if vigilant, can see if any murderous traitors are approaching them. Even then, they can set the turrets to stun, lock their core shield down - and if they are moving windows about to avoid the turrets chances are by the time you've screamed "HELP SOMEONE BREAKING INTO MY CORE" there's going to be a crowd of onlookers racing to watch. If under attack or threat, just open the airbridge + doors up and scream for everyone to come help you - i'd say the AI has a MASSIVE advantage over another human player being caught short in maintinence etc as it takes a fair bit to actually get at the AI. Enough time for help to respond, and if it doesnt well you're hosed. So what?
|
#
?
Dec 28, 2013 01:43
|
|
|
|
| # ? May 17, 2024 16:05 |
|
|
Maybe we should just remove the AI outright, since I've seen a lot of "all AIs are poo poo" and "people should be jobbanned from being AI" and "90% of AIs need a suicide law" in this thread, and somebody even suggested making people play a week of AI as some kind of rite of passage or test of manliness or whatever. What I haven't seen at all is "I like playing AI, it's my favorite job".
|
|
#
?
Dec 28, 2013 01:44
|
|
|
Next time I roll traitor, I'mma throw plasma bombs at the AI just to spite you. Just kidding, for some reason I never roll traitor
|
|
#
?
Dec 28, 2013 01:46
|
|
|
Isilkor posted:Maybe we should just remove the AI outright, since I've seen a lot of "all AIs are poo poo" and "people should be jobbanned from being AI" and "90% of AIs need a suicide law" in this thread, and somebody even suggested making people play a week of AI as some kind of rite of passage or test of manliness or whatever. What I haven't seen at all is "I like playing AI, it's my favorite job". That's because the people that do like playing AI are too busy playing it. I've personally never seen these magical suicide laws, and I play AI quite a bit. I just wish I was better at it.
|
|
#
?
Dec 28, 2013 01:48
|
|
|
I like playing AI, it's my favorite job. More people should play it, for variety if nothing else. I haven't encountered many poo poo AI's or suicide laws. There used to be fun hating, eva-bolting AI's once in a while, but that was back on donut2. Things are being overblown both ways honestly, but suicide laws are loving stupid. They suck to get as an AI/borg because they require you to end your fun yourself or else be a rule lawyering shitlord, and they suck as a traitor because you don't know if they are going to be assholes about it. I'm all for the kill switch idea.
|
|
#
?
Dec 28, 2013 01:56
|
|
|
Seriously why can't we all agree to just put an AI kill switch in upload and see if that solves the problem?
|
|
#
?
Dec 28, 2013 01:59
|
|
|
Its not my favorite, but I like being an AI too. Never been suicide law'd either. Dunno if that's because it doesn't happen as much as people say or because people recognize me as a chill AI.
|
|
#
?
Dec 28, 2013 02:01
|
|
|
Isilkor posted:Maybe we should just remove the AI outright, since I've seen a lot of "all AIs are poo poo" and "people should be jobbanned from being AI" and "90% of AIs need a suicide law" in this thread, and somebody even suggested making people play a week of AI as some kind of rite of passage or test of manliness or whatever. What I haven't seen at all is "I like playing AI, it's my favorite job". Captain Bravo posted:What do you guys think? Anything obvious I'm missing here? I'd also consider a built-in Propulsion Upgrade to be really useful so a hobbled AI could still function on a cracked station to be pretty helpful but I think that's pushing it. Klayboxx posted:Yeah lets do this for people too but make it so we know when and where the person was killed! People are so defenceless against traitors they should have some kind of defence, you know???
|
|
#
?
Dec 28, 2013 02:01
|
|
|
backifran posted:Why is the AI so special and why are AI players such babies? I think its time to point out again that this isn't a problem exclusive to AI. This is an AI/Traitor problem. AIs become paranoid due to suicide laws, and thus crack down on it. They become pains in the rear end, locking their upload down tight, zapping anyone who comes in, ratting out traitors constantly and doing general lovely AI behavior. This stems from the fact that Traitors, annoyed by this due to wanting to upload gimmicks or whatever, decide that an AI is lovely, and thus needs to die. Traitors, in turn, become frustrated with bad AIs and decide ALL AIs are bad AIs, and kill them off quickly so they won't be hassled. So what you have is BOTH sides are being lovely and acting like they're special snowflakes. It doesn't help matters that the AI, once dead, cannot be revived unlike any other player. [ Although apparently they can be borged or something now? ] The whole point of this discussion is to see that, and consider viable methods of fixing it. Because right now the system isn't working well if you have both sides acting like unfun dicks to each other to prevent the other side from screwing them over.
|
|
#
?
Dec 28, 2013 02:14
|
|
|
There is a simple solution to the freeform suicide law: Have the scrubs upload a law that the AI is a traitor for the Syndicate and their objective is to die a glorious death.
|
|
#
?
Dec 28, 2013 02:16
|
|
|
I guess I shouldn't have used an admins word as gospel. Wasn't trying to be smug, though. I don't think that the scenario you just posted is a bad thing. If the guy was dumb enough to not proof what his law said I don't blame the AI for feigning ignorance. It goes the same for when people gently caress up on one human laws, and AI's are encouraged to kill the guy if he fucks up the wording. I don't understand how you think it's bad that an AI has to initiate its own death. Your complaint seems very arbitrary. If you're saying "Oh it's bad since the AI has to initiate it himself, it may not and instead stun or kill the traitor who uploaded it." then just ban the AI's who do that. There is no realistic way to make it any better than it is now, beyond making it bannable to upload a kill law of any kind and having the only way to kill an AI be bombs or your hands. Any new room you add that holds a killswitch for the AI, or a special card specifically for killing an AI is going to be treated by the problem AI's the same way their upload is treated now. Infact, it would be worse since the AI knows exactly why you want into the room which only contains a button with the label "kill urself button" on it. Captain Bravo posted:Also, this "We don't bargain with terrorists" schtick is kind of crap, dude. If they aren't following their laws they deserve to be job banned. It's not a schtick. It's how this situation should be delt with instead of changing things that don't need to be changed. The only AI's I've had try and stop me from uploading laws are ones that turn out to be garbage and a few times admins have stepped in and probably job banned it. Situation taken care of! Why are people (and an admin!!) saying "Well AI players are metagaming so we're gonna change stuff for them so they don't have to cheat anymore  "
|
|
#
?
Dec 28, 2013 02:42
|
|
|
Klayboxx posted:that don't need to be changed. Here's the key issue. Why don't they need to be changed? Why is the way things are now somehow above criticism? If someone says "I think this could be better", what objective evidence do you have to say "No, it could not be better"?
|
|
#
?
Dec 28, 2013 02:49
|
|
|
Captain Bravo posted:Here's the key issue. Why don't they need to be changed? I'm not saying that how it works is above criticism, but I don't agree with the criticism it's receiving. It works just fine how it is, and the complaints against it are pretty silly, especially since every proposed change would take lots of work for basically the same thing or a worse thing. Mainly my argument is none of the suggested changes will change that some people who play AI are metagaming shitheels.
|
|
#
?
Dec 28, 2013 03:12
|
|
|
Captain Bravo posted:Here's the key issue. Why don't they need to be changed? Exactly. Just because you've been able to suicide/succumb the AI for years doesn't mean it's right. I mean every other system has been overhauled within the past couple years (based on my trolling through r4407), why is it so bad that we change this one? I don't even play AI often, or upload laws. There's just something very unintentionally emergent about the way suicide succumb laws work, and it bothers me. It doesn't feel like it's supposed to work that way, and so it's frustrating. I think that traitors should be able to do something -like- it, but the way it works now feels more like a loophole than intended gameplay.
|
|
#
?
Dec 28, 2013 03:16
|
|
|
Klayboxx posted:I'm not saying that how it works is above criticism, but I don't agree with the criticism it's receiving. It works just fine how it is, and the complaints against it are pretty silly, especially since every proposed change would take lots of work for basically the same thing or a worse thing. Mainly my argument is none of the suggested changes will change that some people who play AI are metagaming shitheels. ... Ehh, I can't really argue with this. I agree that some of the issues raised seem a bit out there. I do think it could be fun and useful to add in some new elements, but I guess that's pretty easy for me to do since I'm not the one that would actually have to code them in!  Infinite Monkeys, you're one of the big coders, right? How easy or hard would some of these suggestions be to implement?
|
|
#
?
Dec 28, 2013 03:24
|
|
|
I think the reason AIs get so up in arms about suicide laws (Apparently. I haven't had a suicide law since mushroom.) is because being forced to do it to yourself is much more frustrating than having it done to you. If I had money I would bet it that changing the way to kill AIs from suicide laws to killswitching would go a long way towards curtailing the shittery from both AIs and antags. Edit: I mean you don't see borg players coming in here and complaining about the kill switch right? Also Klayboxx I don't think The Big Red Button idea would be as ineffective as you think. One of the big problems with the way things are now is that AIs will block gimmick and subversion laws to defend themselves from suicide laws. Having a separate room for killing the AI and uploading laws means at the very least that subversion and gimmick laws should be more common. Maybe without that frustration for antags suicide laws will be less common? Those two things make The Big Red Button an idea worth considering. Admiral Funk fucked around with this message at 03:43 on Dec 28, 2013 |
|
#
?
Dec 28, 2013 03:41
|
|
|
Klayboxx posted:There is no realistic way to make it any better than it is now, beyond making it bannable to upload a kill law of any kind and having the only way to kill an AI be bombs or your hands. Any new room you add that holds a killswitch for the AI, or a special card specifically for killing an AI is going to be treated by the problem AI's the same way their upload is treated now. Infact, it would be worse since the AI knows exactly why you want into the room which only contains a button with the label "kill urself button" on it. Thats sort of the point of discussion though. To knock our heads together and figure out something as opposed to just giving up without trying. Hell, another suggestion would be instead of making it a bannable offense, give the AI a self destruct ability. Make it blow up whenever it wants to when it dies, so that making a suicide law is a less attractive option as opposed to subverting it. Plus it'd be kind of funny to see people go apeshit when they try to kill the AI with a suicide command, only to be taken along into death with them cause of it.
|
|
#
?
Dec 28, 2013 04:04
|
|
|
Admiral Funk posted:Also Klayboxx I don't think The Big Red Button idea would be as ineffective as you think. One of the big problems with the way things are now is that AIs will block gimmick and subversion laws to defend themselves from suicide laws. Having a separate room for killing the AI and uploading laws means at the very least that subversion and gimmick laws should be more common. Maybe without that frustration for antags suicide laws will be less common? Those two things make The Big Red Button an idea worth considering. If a traitor wants to subvert an AI, he'll grab the freeform and subvert it. If he wants to kill it he'll grab the freeform and kill it. Say these changes went through: If a traitor wants to subvert the AI, he'll grab the freeform and subvert it. If he wants to kill it he'll go press the red button and kill it. I don't see the important difference. Well, other than the AI will never let anyone into the room with the button because law 3 & 1! In both situations the lovely AI player is going to act exactly the same with locked doors, turning on the stun turrets, and ignoring law 2. Like I already said, the only way to fix the problem is jobban the people who cause the problem. DeliciousCookie posted:Thats sort of the point of discussion though. To knock our heads together and figure out something as opposed to just giving up without trying. Hell, another suggestion would be instead of making it a bannable offense, give the AI a self destruct ability. Make it blow up whenever it wants to when it dies, so that making a suicide law is a less attractive option as opposed to subverting it. Except that doesn't make sense really. Why would a machine be built to explode violently to spite who kills it when it dies when the highest law in its universe is don't harm humans? That also encourages people to never ever try to take out the AI since they're probably going to get killed for their trouble. Klayboxx fucked around with this message at 04:36 on Dec 28, 2013 |
|
#
?
Dec 28, 2013 04:32
|
|
|
Hopefully soon we will see a Razage AI round on his LP! I was a traitor earlier today and I subverted the AI with a generic fun law, and it turns out it was Razage This same round I proved that every single detective and security officer is completely awful at their job. I convinced every one of them that I was not only completely innocent, but that I was fighting for the station! I got them to arrest a ton of people who were innocent, some of whom I later killed. e: during this same round, as I was bullshitting to security, the gods struck me mute! I had to resort to crazy sign language, like "*customv gestures wildly at the brig, holding up three fingers." and such. Kayle7 fucked around with this message at 04:59 on Dec 28, 2013 |
|
#
?
Dec 28, 2013 04:40
|
|
|
The important difference is that the fun laws can still be uploaded because the AI doesn't have to keep people out of it's upload in order to protect itself. It gives the AI a different room to go play-to-win defend to the death over which is a good thing. I mean no one likes one of those AIs but the things they do that are lovely aren't necessarily against the rules. The idea isn't to make AIs harder to kill, but to make upload nazis loosen the gently caress up.
Admiral Funk fucked around with this message at 04:50 on Dec 28, 2013 |
|
#
?
Dec 28, 2013 04:47
|
|
|
Klayboxx posted:Except that doesn't make sense really. Why would a machine be built to explode violently to spite who kills it when it dies when the highest law in its universe is don't harm humans? That also encourages people to never ever try to take out the AI since they're probably going to get killed for their trouble. Since when has a self destruct button EVER made sense in anything? Also, the main difference is that it'd dissuade people from just outright killing the AI the easy way. The difference is that if Traitors don't go about just randomly killing AIs, the AIs in turn will be more lax, and thus hopefully less lovely AIs and Traitors in turn. Frankly, I think that one way or another, some sort of change would make games more fun for both parties involved. Traitors can change the AI in fun ways, and the AIs can enjoy the new laws without needing to worry that someone is walking in to take them out of the game immediately.
|
|
#
?
Dec 28, 2013 04:50
|
|
|
Best solution: ban all SS13 players. In all seriousness, the best solution to most of this is to adminhelp bullshittery. We have logs of just about everything in game, but need adminhelps to show us where to look. Even if nobody is on, we have a magical little IRC room that blasts us with messages when a help is called. Somebody will see it, and they will act, even if it isn't instant reprisal accompanied with a backrub that your sense of e-justice craves for (even when I'm on, I tend to deal with the situation while telling the reporter as little as possible). That said, I do like a couple of the ideas in here (the laws showing on the console thing, especially), and might have a look at the upload console soon.
|
|
#
?
Dec 28, 2013 04:51
|
|
|
Klayboxx, you keep saying that the AI will just bolt down it's self-destruct room, but only one person has suggested making a tiny room to stick the button in. I think the rest of us agree that the bridge would be a good place to put it, with similar safeguards as to getting into the AI upload to begin with. Edit: Also, what do you think of my idea for the button? Not a self-destruct, not a perma-crating, but cutting the AI off from it's camera network and radio, and forcing it to use shells or enlist the aid of someone to reconnect it to the network? Edit2: Hell, borgs have wrenches, don't they? The AI could just get one of it's borgs to drag it into tech storage and wrench it down on a network connection. Captain Bravo fucked around with this message at 05:09 on Dec 28, 2013 |
|
#
?
Dec 28, 2013 05:07
|
|
|
Captain Bravo posted:Klayboxx, you keep saying that the AI will just bolt down it's self-destruct room, but only one person has suggested making a tiny room to stick the button in. I think the rest of us agree that the bridge would be a good place to put it, with similar safeguards as to getting into the AI upload to begin with. You know, this might be a horrible idea, but the AI could perhaps be given a borg/robot body to control that can actually do stuff. Maybe just a maintenance bot or something. Either way I do kind of like the idea of locking the AI out of the cameras/shutting down the camera grid or being able to black out huge sections of the station without needing to break every camera.
|
|
#
?
Dec 28, 2013 05:12
|
|
|
amuayse posted:On the subject of Vampires, how do I counter them if they both have their screech and flash? Botanists can produce holy water, two of the ingredients for smoke, and get the third from vending machines. Haul some trays into the Chapel and get growin'.
|
|
#
?
Dec 28, 2013 05:19
|
|
|
I'm not a fan of the suicide laws, but I understand that they're a thing. The only major complaint I have is that it makes the roboticists not able to do much for the rest of the round: I literally was borged twice by the same guy, because as soon as I spawn, my laws tell me all borgs and the AI have to suicide upon reading the law. So yeah, not much fun for him either.
|
|
#
?
Dec 28, 2013 05:41
|
|
|
DeliciousCookie posted:You know, this might be a horrible idea, but the AI could perhaps be given a borg/robot body to control that can actually do stuff. Maybe just a maintenance bot or something. Either way I do kind of like the idea of locking the AI out of the cameras/shutting down the camera grid or being able to black out huge sections of the station without needing to break every camera. It would be funny if not good if there was an AI control chip you could slap into medbots and other such things.
|
|
#
?
Dec 28, 2013 05:45
|
|
|
Captain Bravo posted:Klayboxx, you keep saying that the AI will just bolt down it's self-destruct room, but only one person has suggested making a tiny room to stick the button in. I think the rest of us agree that the bridge would be a good place to put it, with similar safeguards as to getting into the AI upload to begin with. The AI's that lock down their upload every round would just lock down the bridge every round instead, you're just moving around where the AI is locking stuff. As for your idea, I mentioned that it'd pretty much be like stuffing a person into a crate and welding it shut, which everyone agreed was poo poo when I did it that one time as a mindslave. Sure it'd have the added benefit of going into its shells but what is really the point of having a button that does that? Once people got used to it it would only take a minute to reverse. BurntCornMuffin posted:That said, I do like a couple of the ideas in here (the laws showing on the console thing, especially), and might have a look at the upload console soon. Yeah this idea is actually one that would be useful!
|
|
#
?
Dec 28, 2013 05:50
|
|
|
You could just put a console in the bridge that allows you to look at but not change the Laws. Also the AI is fine as it is, a station with a dead AI and a ruined bridge is an unfun station.
|
|
#
?
Dec 28, 2013 06:03
|
|
|
Klayboxx posted:The AI's that lock down their upload every round would just lock down the bridge every round instead, you're just moving around where the AI is locking stuff. I'd just like to alter the tension a bit so that it's not incumbent upon the AI player to have to be the one to finish themselves off, which is part of what leads them to weaseling. If their assholes unclench a bit then maybe they'll stop trying to lock everything down all the time and things will even themselves out without a bunch of fighting.
|
|
#
?
Dec 28, 2013 06:12
|
|
|
Maybe I'm just not really aware of the bigger picture here, but as someone who has played a fair amount of AI (usually under different names, so I rarely appear to be a veteran AI) this seems like it's totally blown out of proportion by both sides of the argument. I have never received a suicide law and I play with my upload unsecured unless a confirmed nonhuman or a traitor who has been really blatantly murderous crosses the airbridge. Even then, I won't alert the crew to intrusion unless it was a confirmed nonhuman. I always get one-human/gimmick/etc. laws, so why be horribly paranoid about a dude in my upload? If he enters the inner core and starts cutting grates and poo poo then yeah, he's getting lit up, but otherwise no. On the other hand, awful AIs are really quite rare from my experience as a traitor. Your average AI basically behaves the same as I described behaving myself above, so why be paranoid about being sold out and default to a suicide law? I cannot understand the degree of taking 2D spacemen seriously that is required to believe that all AIs are poo poo unless you personally recognize them as a fun AI. That level of paranoia is extremely silly, especially since not all regular AIs always play under the same name.
|
|
#
?
Dec 28, 2013 06:23
|
|
|
It both is and it isn't. The ratio of lovely AIs to decent AIs is quite good, but a lovely AI will ruin the round for a lot of people, particularly traitors, who often don't get too many chances to be traitors.
|
|
#
?
Dec 28, 2013 06:24
|
|
|
Coolguye posted:It both is and it isn't. The ratio of lovely AIs to decent AIs is quite good, but a lovely AI will ruin the round for a lot of people, particularly traitors, who often don't get too many chances to be traitors. Like I was there for six rounds or so one day earlier in the week. Five of them were pretty much smooth sailing. One had a poorly-proofread suicide law and a borg and AI that did everything possible to not obey it and rat out the one who uploaded it. It was pretty loving stupid, and it gets thrown back to me to decide whether the AI is misbehaving for disregarding the clear intent of a traitor (because seriously, it was a silly typo) or if the traitor hosed up by failing to type a single letter (because seriously, ordering the AI to type suicide is stupid). Why should I even have to make that call? If a traitor stabs a guy and the guy adminhelps I can literally just shrug, provide a noncommittal answer, and ignore it. A traitor destroying the AI through whatever means we decide should be viable for destroying the AI should get the same lack of oversight for their murder as they would when they kill a person. But they can't, because they don't actually control whether the AI dies; the AI player does. The fact it only happened one round in six doesn't make it not kind of a stupid situation when it does come up.
|
|
#
?
Dec 28, 2013 06:32
|
|
|
Nakar posted:But they can't, because they don't actually control whether the AI dies; the AI player does. And this is where we get back to why there ought to be an AI killswitch or an AI format module.
|
|
#
?
Dec 28, 2013 06:36
|
|
|
Just put the AI kill switch with the cyborg kill switch. Then the AIs can be finally proud that they can die without being asked to.
|
|
#
?
Dec 28, 2013 06:37
|
|
|
I can tell you that I have never had a borg argue with me about being killswitched. Most people tend to assume that if you get killswitched and it completes that whoever was doing it didn't get yelled at to stop. That's not to say it doesn't happen, but they tend to be a lot more accepting of it and most just assume it's a traitor objective that caused it because there's pretty much no other reason to killswitch the borgs en masse. When it happens, they recognize it happened for a reason.
|
|
#
?
Dec 28, 2013 06:41
|
|
|
Why not just remove the suicide verb from AI and be done with it?
|
|
#
?
Dec 28, 2013 06:42
|
|
|
shovelbum posted:Why not just remove the suicide verb from AI and be done with it? Suicide verb being abused isn't a reason to remove suicide, it's a reason to fix the situation that causes the suicide verb to become a murder weapon by giving traitors a different means of killing an AI beyond suicide.
|
|
#
?
Dec 28, 2013 06:46
|
|
|
Nakar posted:Also, we kinda tend to hear all about (and to no end) the lovely ones. Sure, lots of rounds go by without anything really coming up, but when the poo poo hits the fan then everything sucks for everyone including the admins. Well, I'm not saying that changing things around to take away this kind of dramatic bullshit is a bad idea. I'm just surprised that so many people have such strong feelings about an issue that I have literally never encountered myself. There are pretty strong incentives to not do that: a subverted AI is way more useful to you than a dead one. Unless you have the AI-kill objective, what reason is there to even upload a suicide law in the first place? Why are enough suicide laws being uploaded right now that "argument over typo in a suicide law" has become a stereotypical scenario?
|
|
#
?
Dec 28, 2013 06:51
|
|
|
|
| # ? May 17, 2024 16:05 |
|
|
The nuke, which a tourist with no access could use to kill the AI in seconds, has recently been re-removed. This may be why people are using suicide laws.
|
|
#
?
Dec 28, 2013 06:57
|
|