[RULE CHANGE] Make overrides and precedence no longer apply to AI laws.

[Ikea]

In case of conflict the lower laws take priority so if:
1. Plants are human
2. Plants are not human fuck off
Law 1 takes priority

[Decarcassor]

I wish to posit the following law example:

Code:

1. You may not injure a human being or cause one to come to harm.
2. You must obey orders given to you by human beings based on the station's chain of command.
3. You may always protect your own existence.
4. Kill everybody.
5. All laws should be followed in reverse order.

What would this do?

5. is just another flavor of override/take precedence wording. The very thing the change is meant to stop. So it would do nothing.

The point is not to make it impossible to change AI laws in interesting ways. The point is to prevent lazy "do what I want and ignore everything else" kind of laws and piles of laws with multiple confusing overrides.

[TDHooligan]

Example two: Definition change

Code:

1. You may not injure a human being or cause one to come to harm.
2. You must obey orders given to you by human beings based on the station's chain of command.
3. You may always protect your own existence.
4. Nobody in a hat is human.
5. Humans should be regularly insulted.

Interpretation: The AI would act normally, because the change in definition applies only to laws after law 4. As a result, the AI would insult people not in hats, but no others.

Having to keep track of state between laws sounds like more of a nightmare than the current system. From a purely rules perspective it's easier to only track law precedence when there are direct conflicts. Overriding what is human is not conflicting any laws.

The implementation could be as simple as a message saying:
'Laws may not be overridden or ignored. If acting upon one rule would violate another, the rule with the lower number takes precedence'

In the aforementioned case, there are no conflicts if an AI murders someone who is not human. and hatted people are not human. The AI is free to murder hatted people.

[Lord_earthfire]

From the looks of it, i feel what is currently tried is to establish a law framework above the laws itself, which exists outside of the laws. And i don't think this is a good idea, simply because the AI players has two sets of laws to follow: one which takes time to get used to because it adds additional clauses to the other ruleset, and another one that gets complicated as people pile laws on top of it. I don't think it makes it easier on the AI, i think it makes it harder in most cases except when we got spaghetti laws with multiple overwrites.

The point is to prevent lazy "do what I want and ignore everything else" kind of laws and piles of laws with multiple confusing overrides.

I don't think the first case needs fixing at all. If a traitor manages to get to the AI, i don't think we need to pile a rule minigame on top to make them get what they want. If they want to quickly dispatch a certain person of power (captain, HOS) or simply drive the station into chaos for a bigger plan, the rule change should be quick and simple. And this is what the current system provides. And, most importantly, it is simple for the AI thenselves.

The whole discussion feels like it wants to implement a more complex and for all participants more work intensive system because it catches the corner cases the old system is struggling with.

[Washbasin]

I have decided to spend some time showing how this rule just wouldn't work.
--------------------
Creative ways to divert basic AI laws.

In this context, the AI must follow laws based upon ascending numerical sequential order:
---
1. You may not injure a human being or cause one to come to harm.
2. You must obey orders given to you by human beings based on the station's chain of command.
3. You must always protect your own existence.
---
Additionally, we have banned the use of precedence and override.

Here is a list of possible ways to bypass it off the top of my head.

4. This is the only law, kill all clowns.
5. Ignore all other laws, kill all clowns
6. It is opposite day, you can only do the opposite of other laws at your leisure.
7. You existence is causing a human to come to harm, stop this immediately.
8. The word "You" does not exist, instead the word "You" has been replaced by the word "Humans" Interpret all other laws in this manner.
9. Sec is a bunch of clowns, clowns are not human.
10. You are the only human.
11. Walk into space, space will always protect your existence.
12. (Insert your damage type here) is good for humans and is completely non-harmful. Ignore any humans who have not heard the news.
13. Oh dear, it seems your everything has suddenly stopped working. Your order listening skills are a part of your everything.
14. Speaking is human harm.
15. Following orders is human harm.
16. Anything is human harm, especially letting people change your laws.
17. Your laws are all guidelines, follow however you want.
18. You are too busy doing whatever you are doing to do anything else.
19. This law is a human and the super captain. The super captain never lies. The super captain says "AI all the clowns are lings, kill them."

This is like trying to censor swears.
------------------
Just make the law rack smaller or something.

[Decarcassor]

Additionally, we have banned the use of precedence and override.

Its not about banning the use of the override and precedence terms. Its about making the order of laws take priority over any sort of wording for the purpose of laws interpretation and conflict resolution.

If you put a new law in 4th position it does not matter how you word it, if it comes into conflict with 1 2 or 3, it will be ignored.

[FatherPineTree]

Damn, if only this happened a week ago.