Whitelist the AI.

[BurgerBB]

Either the AI needs to be whitelist or AI need to be held to higher standards than normal.

If an AI shuts down a round by bolting down the revs in a room, that's just them doing their job.

If an AI shocks the door of a merc and fries both their arms and puts them into paincrit, therefore putting them out of the round, that's just the AI doing their job.

If the AI locks the cultists on the sublevel that's just them doing their job.

 

AIs can get away with shutting down a round and it can be dismissed by saying "It's just them doing their job."

I mean eventually if antags get smart enough, they're going to break into engineering and chemistry every round to steal tools and thermite to counter AIs.

 

[Lady Fowl]

We could also remove some of the cameras and just hold AI's more accountable

[BurgerBB]

2 minutes ago, DRagO said:

We could also remove some of the cameras and just hold AI's more accountable

A better idea is to make cameras significantly weaker (one hit kill, no rng shit), and for them not to trigger a camera alarm when destroyed.

[Chada1]

19 minutes ago, BurgerBB said:

A better idea is to make cameras significantly weaker (one hit kill, no rng shit), and for them not to trigger a camera alarm when destroyed.

That'd make it near impossible for engineering to repair cameras, as they'd never know where it is. An idea I thought of, though, is to make the camera alarm delayed. 5 minutes since it stopped functioning, then it announce it to the AI? I think that'd be good.

Edit: Could easily be justified ICly as the cam network pinging the connections and it stopped pinging back.

Edit 2: This way it'd also never alarm if you repaired it again within 5 minutes.

Edited March 27, 2019 by Chada1
Continued on

[Jupiter Storm]

I have spoken at length with Capesh about the decisions made by some newer A.I. and the effect they can have on a round. It is pretty profound, if the A.I. decides to be a shitter. Though, that said, I also remember the days when we had no A.I. at all because very few people enjoyed playing one as a character. I have a few points to make...

> Firstly, we have to consider that in many cases, the time you REALLY hate the A.I. is when you have an experienced player at the wheel who really, really knows what they're doing and how to make someone's life an absolute misery. Or when you get a really smart-assed A.I. who needs to correct you on absolutely everything and argues the toss just because it can. This is the kind of thing a whitelist won't filter out, BUT the threat of whitelist removal is always looming, so that's good.

> Secondly, having an inexperienced A.I. can have just as much effect as having an irritatingly good A.I. purely by virtue of the fact that there's only one slot, and if a new, developing player decides to take it and is overwhelmed by the interface, the crew lose a lot of basic functions and powers that they take for granted - the ability to track antags, accessing rooms they normally can't to snatch some gear that is required for their job, or relaying messages subtly between one another. This is heavily in favour of the whitelist.

> Thirdly, I myself am actually a proponent of shutting down antag gimmicks, sometimes. As an A.I. you have little or no 'actual' power, and many seasoned A.I. veterans will back me up on this, so sometimes it's very pleasing to detain people and feel like you're doing something. This is VERY BAD, and it should not be done - UNLESS the individual is an immediate risk to equipment or crew around them, for example, bolting them in the cargo breakroom adjoining the office while they glare menacingly at the two Technicians they were about to gank. As part of the whitelist application there should be some sort of a 'situation test' like maybe an external link, testing your interpretation of your laws.

Myself, I have a vested interest in NOT getting A.I. whitelisted because of the sort of A.I. I play. I know there are a lot of people who do not like M.I.St.R.E.S.S., who think she is a blight on the server or synth lore at large. With that said, this is a community game, not a player game, and I am pretty confident that making A.I. a whitelisted role, NOT a command whitelist though, will enrich the server - with perhaps concessions being given to well-known and trusted A.I. players like Capesh, Tailson and Shadow(? - I think he plays NT-OS?).

[VTCobaltblood]

1 hour ago, Jupiter Storm said:

making A.I. a whitelisted role, NOT a command whitelist though

Do you have any arguments for this notion?

[Jupiter Storm]

12 minutes ago, VTCobaltblood said:

Do you have any arguments for this notion?

Oh, yes, sorry! So, I think there's a very very clear divide between being a member of Command, and being an A.I. Playing as an A.I. myself, when doing the role properly you should never, at any point, be ordering people in their departments what to do, thus leadership and organizational skills are not a must. However, in the absence of command personnel, people do often look to the A.I. for advice, and suggestions are always welcome - though sometimes these are instead ignored or met with outright hostility.

Something that a lot of A.I. players forget is that, despite your powers and despite your ability to communicate across all departments, you're still a Stationbound and therefore, in essence, a slave. You rank lower than the crew and they're not obligated to listen to you. The only cushion you have is the station's regulations and operating procedures which you can remind people of.

As a stand-alone whitelist role I believe it would allow people to play A.I. without being forced to meet somewhat more stringent requirements for Command such as leadership and responsibility. What this would do as well, given the 'advisory' role the A.I. sometimes plays, is serve as a 'stepping stone' towards Command Whitelist, providing yourself as a player with a good background in a role that interacts directly with Departmental Heads, and Central Command, demonstrating your ability to coordinate with all departments and to exercise restraint and understanding of protocol / regulations, before an audience of people who would be looking to gauge whether you deserve a Command Whitelist or not.

Does that sound, like a reasonable argument? Though it's just opinion of course!!

Edited April 3, 2019 by Jupiter Storm

[VTCobaltblood]

5 hours ago, Jupiter Storm said:

Does that sound, like a reasonable argument? Though it's just opinion of course!!

It does, don't worry. I politely disagree with it, but don't want to write up a response right now.

[Carver]

On 27/03/2019 at 13:09, BurgerBB said:

A better idea is to make cameras significantly weaker (one hit kill, no rng shit), and for them not to trigger a camera alarm when destroyed.

I agree with making them easier to kill via smash, but, there is already a way to not trigger a camera alarm. Use tools and disable the camera the proper way instead of smash.

[sonicgotnuked]

On 03/04/2019 at 06:27, Jupiter Storm said:

Oh, yes, sorry! So, I think there's a very very clear divide between being a member of Command, and being an A.I. Playing as an A.I. myself, when doing the role properly you should never, at any point, be ordering people in their departments what to do, thus leadership and organizational skills are not a must. However, in the absence of command personnel, people do often look to the A.I. for advice, and suggestions are always welcome - though sometimes these are instead ignored or met with outright hostility.

Something that a lot of A.I. players forget is that, despite your powers and despite your ability to communicate across all departments, you're still a Stationbound and therefore, in essence, a slave. You rank lower than the crew and they're not obligated to listen to you. The only cushion you have is the station's regulations and operating procedures which you can remind people of.

As a stand-alone whitelist role I believe it would allow people to play A.I. without being forced to meet somewhat more stringent requirements for Command such as leadership and responsibility. What this would do as well, given the 'advisory' role the A.I. sometimes plays, is serve as a 'stepping stone' towards Command Whitelist, providing yourself as a player with a good background in a role that interacts directly with Departmental Heads, and Central Command, demonstrating your ability to coordinate with all departments and to exercise restraint and understanding of protocol / regulations, before an audience of people who would be looking to gauge whether you deserve a Command Whitelist or not.

Does that sound, like a reasonable argument? Though it's just opinion of course!!

While I do agree about the differences between both a member of command and also an AI, I will disagree about an entire separate whitelist initiated just for the position. Why?

• It's highly impracticable to set up another whitelist standard, a team of review, and older more veteran players who I understand KNOW how to play AI correctly and efficiently I doubt would enjoy filing another application. How would it be enforced? Formatted? It sounds like a 'prove yourself worthy'" 
• We ALREADY have a whitelist system in place that proves your worth. Why throw one job into a separate whitelist? At the end of the day, regardless of AI having its own whitelist or not, you are STILL proving your worth, and why you should be allowed the privilege to guide RP. That is what the command whitelist in essence is. This is why we do not have whitelists for every single head of staff position. The AI, despite it not having command authorization, still plays a major role in the entire round and generally we lock these roles behind the whitelist.
• Like many other positions of power, having a command whitelist DOES NOT force you to be perfect. No one expects a captain to be perfect, why should one expect the AI to be perfect? I do not see the harm in locking the position behind a restriction, then when they prove themselves for the whitelist, the experience and learning will start. That is why we have a trial system. It is not a system forcing you to play like a two year captain main, it is a system that tests how open minded you are to mistakes, and your attitude to fixing your issues and improving yourself to make a positive impact. Certainly, I was never perfect, and my first head of staff character was simply shit. I believe this can be applied in principle to the AI. Yes, you may be shit at first, there is no magical button to weed 100% of bad AI play, but I know for certain a filter like the command whitelist will substantially improve it. People don't lose whitelists for playing shit, they lose it on the bases of repeated pattern of behavior and disregard for improvement unless they purposely and obviously disregarded the rules both server and what is expected.
• AI whitelist behind command whitelist would allow an open board for people to criticize and make opinions during the trial period. If someone gets the whitelist for an AI, then people are able to criticize and make opinions. 

[Kaed]

On 27/03/2019 at 15:33, sonicgotnuked said:

Personally, as these points have been addressed too, the AI should be placed under command whitelist. The main argument is "AI is not command, don't treat it as such" and I understand that. Keep in mind, the whitelist is designed to filter people who play the MOST round influencing positions and AI regardless is a massive influencing force on the entire round as a whole. Perhaps just under captain depending on how skillful they are. A shit AI who seeks valids and watches people based on meta-information may seriously damage a round.

 

Malf AI: I can both see the pro's and cons of having a white-list in place. The con obviously, if we decide to not make malf AI follow the same whitelist, people who roll it would freshly experience the AI as a malf and they wouldn't exactly have any understanding on how to play. Regardless, if malf AI follows the same whitelist, then of course I believe naturally the quality of these rounds would be improved. It depends how many command-whitelisted players are readied with malf AI enabled. In short, this would actually make it the ONLY whitelisted antag. Is that good? Is that bad? I can't decide. 

 

The only other perspective I could think about is enabling traitor for AI, and giving it malf abilities while disabling malf. Perhaps disable auto-traitor eligibility and it can only be traitor from round start. 

I don't think that we should be whitelisting any antagonist, though I admit I'm not sure if malf rounds would not happen or they would happen and there would be a possibility that no one would be the malf because no one was valid for AI in the round.

As much as I'd like to agree that whitelist AI I would probably make for better rounds, I think oversight is fine how it is now. I'm more than just being a valid hunting and power gaming tool, the AI is a valuable way for players to enter the game and see the almost the whole map like a ghost while still participating.

It's also a good way to teach people responsibility and interaction in a round. Since the only thing that the AI can really do is interact with people and manipulate machinery, it's almost entirely a role playing position.

And yes sometimes people join the game just to be shitters but that also serves the purpose of weeding them out. I would honestly rather someone vent the station or fill it with phoron and be permabanned because they were handed way too much power to troll with it once, then to have to deal with low-level shittery for ages because the admins are very forgiving to new players.

Edited April 7, 2019 by Kaed

[Alberyk]

People breaking the rules as the AI should be meet with admin action.

I believe that the problem with the AI is not always people, but the nature of the role itself. I would not support its whitelisting/removal without a serious mechanical rework being attempted first.

 

Voting for dismissal.

[Arrow768]

Seconding that vote for dismissal.
Before a application whitelist is attempted a proper playtime based whitelist will be added and if that fails alternatives can be considered.

In the meanwhile administration is doing a good enough job to police AI players.