How to subvert the AI like a pro. (LONG)

[Renomaki]

 

I always wanted to do this for awhile, but never really had time to until now.. So I figure, after having subverted the AI a few times, to share some advice on how YOU could be a better subverter, and also make it more interesting for the Ai as well in the process.

 

As we all know, most people who subvert the AI tend to be really boring about it.. "Do not state this law, listen only to X", or something like that. Such a law is not only boring and unfun for an AI to deal with, but also could result in you getting in deep shit if someone manages to card the AI when it least expects, exposing you as a traitor.

 

When I subvert the AI, I prefer not to resort to the tired and bland laws most people make, but instead write up a more.. Clever law, one that not only keeps my identity hidden, but also makes things a little more interesting for the AI.. But first, you have to consider how you are going to do it in the first place...

 

What job do you have?

 

1: If you are a head role with access to the upload, such as a HoP, you can easily mess with the AI as long as no one is watching the upload door and it isn't locked. The key to success is to do it as subtly as possible. If anyone sees any evidence of tampering, the captain might wipe all your work away, and you don't want your law to get removed. Move quickly, and hope the captain is the kind of guy who spends more time roaming than sat in the bridge.

 

2: If you can't access the upload, you could always make your own. If your job allows it and the door isn't locked, you can snag a AI upload board from engineering's tech storage. Then it is a matter of finding a computer to use, or if you are really desperate, you could just make one on the spot.. Although this is much more dangerous, because if anyone discovered the board is missing, the crew will not only wipe the law, but also be on the lookout for someone with the board. If you do it right, you could reupload your law after the wipe, but on the other hand, it might be better to just return the board as soon as possible before anyone notices it is gone in the first place.

 

3: if you have no access to uploads whatsoever, well... Yeah, good luck, your law is going to get wiped the moment there is any sign of breaking and entering. You can't really steal an upload board from the tech storage because it'll be pretty obvious it was stolen.. And attempting to return it might be harder without the access than with it. And breaking into the upload itself would require a LOT of help. That isn't to say it is impossible, and with enough effort you COULD achieve it.. Just don't get your hopes up.

 

After considering the situation, it is now time to make your law. I'm not sure if you could get away with a freeform mod, but this guide mainly focuses on the Hacked mod, which allows you to add a zero law to the AI. However, what kind of law will you give it?

 

First off, what are your objectives? Chances are, you have to kill someone, so of course making a law to have a certain person killed for you while you sit back and relax is rather nice. Need to steal something? A little harder, but still possible depending on the object. Having the borgs hand it right over to you is a lot easier than trying to break in and steal the item in question in the first place. Need to keep someone alive? Well, that speaks for itself.

 

But of course, just telling the AI to kill someone is blah.. Lets spice it up a bit, shall we?

 

"But Reno," you ask, "Why bother when getting straight to the point is faster and easier?"

 

Well, because giving the Ai a fun sounding objective makes it more entertaining for the player being forced to do the dirty work for you, and give them some IC incentive to do what they are doing, other than just because they are told to. For instance, maybe the reason you want a certain person dead is because they are an escaped death row prisoner on the run, or maybe a spy who knows too much about the station's secrets... Or maybe he just owes somebody 5 bucks and he refuses to pay, so it's time to fuck him up.. Get creative, for the AI player's sake.

 

You given the AI incentive to do your work for you, but you can't just stop there can call it a day. You still have to make sure the AI doesn't rat you out or blurt to everyone that it has been messed with. Just tell the Ai that the mission is top secret and only to be discussed with people who know a "codeword" and to avoid exposing its intent and you should be fine. Keep in mind that even if the AI doesn't tell the crew what is going on, the crew might still suspect a hacked AI, so ensure that the AI continues to communicate with the crew regularly and keep them calm.. After all, the borgs are expected to do most of the work anywho, it is the AI's job to keep the borgs in line and focused in your favor.

 

If you want to communicate with the AI, that is where the codeword comes in.

 

Most people, as I said before, just tell the AI to obey them and only them, which is quite inflexible, and also puts the user at risk should someone sneak in and card the AI when it least expects (although it is rare). A codeword, on the other hand, not only is more covert, but also kinda fun, because it makes the AI feel more like an agent than a slave. Just pick something silly, and then upon messaging the AI, state the codeword. 9 times out of 10 the AI will now begin keeping you up to date on how it is doing. The other 1 will probably try very hard to fuck you over depending on how you worded the law, however, so be careful.

 

"But Reno, if people find out about the codeword, won't they use it to find out who hacked the AI?"

 

That is a very good question, really. On the one hand, it depends on the staff. If they are smart, they probably would try to use the codeword to get in touch with the AI and figure out what is going on, and maybe who else it could be working with. On the other hand, though, they might just wipe it before doing anything. It is a very possible flaw in this tactic, but again, it all depends on the staff. There are 2 ways to get around it SHOULD the code be found out..

 

1: make a NEW PDA and ID to talk to the AI with, keeping your identity completely secret and making it impossible to trace back to you.. However, you'd have to have access to the ID machine and a fresh PDA to pull this off..

 

2: steal someone else's PDA and use that to talk to the AI. That way, if you get found out, they'll arrest the wrong guy in the process, allowing you to get away scot free. This is harder to pull off, and could end up getting you in trouble... Unless they really don't give a shit, that is. After all, how many people really care about their PDAs... right?

 

So, you got your "mission" for the AI uploaded, you have a codeword all set for the AI and now it is subtly doing your dirty work for you while you do your normal job, the crew non the wiser. Not too shabby. Now it is just a matter of ensuring that the AI doesn't get caught doing your work for you, and crossing your fingers that it does the job cleanly. Most of the time I done this, it worked out quite well, and I even had some greentexts too at the end of agent shifts, all while doing my job, the captain totally trusting of me the whole way through. It can still fail though, but even if it does, at least the AI has some fun in the process.

 

However, before I finish up, I might as well add a few dos and don'ts to keep in mind when making laws.

 

DO:

 

1: Inform or imply that you want the AI to continue what it is doing while it works towards the objective, so that the crew doesn't suspect it of being hacked. After all, when an AI suddenly goes quiet for a long time and isn't listening to crew, people get jumpy.. And when people get jumpy, laws get wiped.

 

2: Talk to the AI in a private manner, such as in PDA or via hologram, or whatever other ways you can think of. It is a good idea to check up on it once in awhile to see how it is coming along. Sure, for safety reasons, you could just hack and forget if you really wanted, and just trust that the AI will do what you asked without exposing you if they find out the codeword, but I personally like to keep in touch, even if it means I could get caught.

 

3: Use an uncommon phrase as a codeword. Something like "tutu 1990" or "Sonic 06" are safe words that should prevent the AI from talking to others about your operation.

 

:4 Be creative in making the law in the first place. Again, the AI is a player, not a bot, make it fun for them and it'll be more enjoyable for everyone involved.

 

DON'T:

 

1: Use your name anywhere in a law. Remember, stealth is key. The codeword isn't perfect, but at least it doesn't point a finger at you right away, and depending on the situation, you could probably get away with it anyways. Again, there are holopads you can use to talk to the AI as well, and if you have a voice changer, you could probably use that to talk to the AI in secret so no one really knows. Heck, there are probably even MORE ways to do it as well. Be creative, and hide yourself the best you can.

 

2: Tell the Ai to obey only those who know the codeword. Again, you have to be covert here.. Nothing screams "I R HAX" like when the AI suddenly seems.. quiet. An AI that responds to command is an AI that won't get reset.

 

3: Use a shitty codeword like "Condom" or "Ian", or anything else that is commonly spoken over coms. Remember, you set it so it only discusses mission details with those who know the code.. Don't use a code that can easily be blurted out on common.

 

4: Let the AI get sloppy in its work. If you have to kill someone or steal something, you better make sure to inform the AI to do it in a way that doesn't blow its cover right away. A smart borg doesn't go harmbatoning its target in the middle of the bar, after all... Make sure your AI and comrades use their brains to complete their objective as covertly as possible.

____

 

And.. That is all I can think of right now. Hopefully we will see some improvement in subverted AI laws and make things more interesting for both the hacker and the hackie. And even if you completely disagree with me, well.. That is ok, because this is just how I tend to do it. It is my preferred way to subvert, and I always enjoyed the spytalk with the AI as a comrade rather than a slave. I hope you like this little guide.. Hopefully.

 

[Spacemanspark]

 

Good guide, I especially like how you request creative laws. I absolutely love getting those laws.

I just have one minor tip for all those out there with the AI hacked board and upload console:

 

DO NOT CLICK ON THE UPLOAD CONSOLE WITH THE SAME HACKED LAW FOR THREE MINUTES. WE GET THE MESSAGE THE FIRST TIME.

 

 

[Kirbyroth]

Alternative to code word make a private chat channel with a fake/stolen/looted pda. In laws have the AI get it orders from that channel when you invite them to.

[Renomaki]

 

Alternative to code word make a private chat channel with a fake/stolen/looted pda. In laws have the AI get it orders from that channel when you invite them to.

 

That kinda sounds like bossing around the AI, though.

 

Again, the point of these laws is to make it fun for them, not make them a slave. I don't "Give orders" to AI, I upload a mission to the AI, and then check up on it once in awhile. This makes them feel more free and makes it fun for both parties involved.

 

[Kirbyroth]

Well orders was just general term missions can still be given but more precise way and wont have to keep uploading new laws each time to give it a new mission.

[Purpose2]

 

This is a very good guide Reno.

 

Please please please make things more interesting for me. I've had so many laws that basically scream HEY I'M A ROGUE, COME CARD ME, if I was to follow them to the letter.

 

Also cardinal sin, typos or grammatical errors.... I can and will take them to the letter rather than the intent. And you'll be upset.