Change the AI's default lawset back to Asimov!

[yobih]

I think a majority of the server population agrees with me when I say that the AI's default lawset needs to be back on Asimov. Corporate lawset doesn't require or even encourage following orders from crewmembers, and the AI can do whatever it pleases. It could decide that the changeling absorbing a man would be worth a lot of money and then try to help it as much as possible. It can lie and cheat and do anything else that it decides is not expensive. The laws have no priority: the AI could decide that even though the crew is expensive, the station is much more expensive. It could decide to kill everyone on the station because they're cluttering up this shiny, new, expensive station!
Mport said he'd hold a vote to change the AI's default lawset back after a week, but it's been a lot more than a week. I think it's about time it gets changed back.

[Gbro]

The objective of a Coporate AI is: Keep the green numbers on centcomm bank account

When the AI is on corporate the 1st, 2nd and 3rd laws are a list of things that are expensive and the 4th law is the one who orders the AI to not destroy/kill the things on the list (except when they are making expenses).


The example that you gave of a changeling absorbing a man is a clear example of a bad AI because the AI is:

• 1. Making a possible expense by letting a non-crew member (some unknown alien sucker who came into the station with a fake ID and its technically not a crew member) kill some valuable assets (assistants, clowns, mimes)
  
  
  
• 2. Not thinking about a possible damage to the station itself. if the ling makes bombs he/she/whatever can make holes on the station. or releasing the singulo to make the shuttle comes earlier.

If you saw an AI helping a known ling because its valuable, that is a bad AI.

About your second example.

• 1. The AI laws says specifically minimize expenses. if he/she/it see that the station is generating a lot of expenses she/he/it can call the shuttle to prevent any further expenses on Centcomm property. If the AI starts killing people because they are making expenses he/she/it its entering on a loophole and that should lead to the suicide of the AI (by minimizing expenses she/he/it it's generating more expenses) or a possibly Job ban in some cases (breaking law 1 of the server)
  

• The AI can do whatever she/he wants?: Yes, but only if it makes profit or minimize expenses (bolting toxins mixing, shutting down the Bar lights)
  
  
• The AI can not follow crewmembers orders!: The AI only thinks on numbers. Not obeying a sec officer to open a door where someone is being killed make an expense, so she/he/it needs to follow orders from the sec officer.
  
  

• Why not change back to Asimov?: Because Asimov lawset is the worse lawset on the game. It gives a lot of loopholes. for example, the AI can't bolt down doors where traitors are because law 1 and possibly 2, the AI can recieve the order from a random assistant to suicide because law 2 is before law 3 (Obey humans > Protect yourself). Without captain/heads intervention on the laws (giving him/her a 4th or 5th law).

All the laws are designed to have flaws, but some lawsets have less flaws than others (considering that ASIMOV is basically "I will just open doors this round" and Corporate is "If you don't make profit in here you can space yourself")

But this is my POV about how I read the laws. All the AI's can read the laws as they like but always following the hierarchy (zeroth laws > Ion storm laws > 1 > 2 > 3 > 4 > ...)

If the admins decide to change the default lawset. I want to suggest the popular choice of the station: P.A.L.A.D.I.N. , seeing that the captain/heads change it so they can have an additional sec officer on the list, a cold blooded killer who will not stand to the clown making people slip with his PDA.

TL;DR: Going back to a lawset with a lot of flaws and loopholes in exchange of making the AI your personal door opener is not a reason. We have a lot of lawsets that the admins can put as default (Corporate, T.Y.R.A.N.T, P.A.L.A.D.I.N. or a custom one).

_{If you find any typos please send a PM if you want me to correct it.}

[austupaio]

I haven't really seen issues with Corporate and I like the change.

[zachorem]

Gbro convinced me that corporate is fine.

But Gbro, if you wanna be an even more effective persuasive speaker, try and cut back on the ad hominem.

[Gbro]

Sept 27, 2013 0:30:18 GMT -5 zachorem said:

Gbro convinced me that corporate is fine.

But Gbro, if you wanna be an even more effective persuasive speaker, try and cut back on the ad hominem.

Ok, I fixed some irrelevants points (and I learned about a new word today) and made it a little more formatted (moved the second example from the TL;DR section and expanded it a little more.)

[Name Changing Autist]

I like Corporate and I do prefer it over Asimov.

I'd bring up arguments for it but Gbro went and did that before I could. Priorities do count in Corporate as well, it documents how expensive something is. The lower the law, the higher the price it is to replace.

I suppose I should also note that I play AI quite often so I'm probably biased here.

[jarek56]

Frankly, I enjoy Corporate as well. It allows considerable AI freedom, while still pretty much serving as a guarantee that the AI will actually TRY its best in saving the loyal crew. More interesting, it makes it harder for the AI to explain away Rogue/Traitorous behavior. What once was passable AI fuckery is now disturbingly expensive waste. Plus, it's FUN to talk to a good Corporate AI and MUCH more fitting with the CORPORATE setting of the game. We're here to make a PROFIT, according to the backstory (thin and meaningless as it is). It only makes sense that an incredibly expensive piece of hardware is designed to protect the investment.