AIs, Honesty, Murder and Trolleys

[Flyndaran]

That conflict would "break" the character regardless of the choice it made. Play that up heavily. Morally justified or not, that character killed innocent people, and only a cold blooded sociopath could handle that easily.

[vicky_molokh]

Originally Posted by Flyndaran

That conflict would "break" the character regardless of the choice it made. Play that up heavily. Morally justified or not, that character killed innocent people, and only a cold blooded sociopath could handle that easily.

This isn't an issue of breaking. The original instance of character responsible for the act (a) is dead and (b) failed to complete the act anyway.
The question is that since Honesty not merely makes difficult, but outright prohibits murder (unlike lesser crimes and many other forms of killing), whether Honesty was hacked away. And the answer to that depends on whether an act of mass killing is classified as murder when it prevents even more deaths under all jurisdictions. (The AI was from EU/France.)

[Flyndaran]

Honesty does have a self control number. This character simply saw a reason to resist it "for the greater good". Memetics was simply how it was convinced of the reason.

[vicky_molokh]

Originally Posted by Flyndaran

Honesty does have a self control number. This character simply saw a reason to resist it "for the greater good". Memetics was simply how it was convinced of the reason.

Notice that it only allows an SC roll to break unreasonable laws; the text clearly states that, and then proceeds to make special mention that one with Honesty may never murder. So the question is whether such an act is always considered murder (and it seems to be a yes, but it would be important if somebody pointed to cases when it is not).

[Anthony]

Originally Posted by vicky_molokh

The question is that since Honesty not merely makes difficult, but outright prohibits murder (unlike lesser crimes and many other forms of killing), whether Honesty was hacked away. And the answer to that depends on whether an act of mass killing is classified as murder when it prevents even more deaths under all jurisdictions.

Murder is unlawful killing, and thus actually depends on the jurisdiction you're in.

[vicky_molokh]

Originally Posted by Anthony

Murder is unlawful killing, and thus actually depends on the jurisdiction you're in.

Which is precisely what I'm interested in - whether a jurisdiction in the EU (particularly France) can have a definition of murder that would not apply to a Trolley Dilemma. Because the NPC in question either had to have Honesty hacked away, or have been convinced that it wasn't committing murder.

[Keiko]

Could it have convinced itself even if only for a breif time? If SAIs are largely human we can do a good job of deluding ourselves particularly for what we see as the greater good.

[vicky_molokh]

There's definitely a delusion involved. The question was whether the delusion was accompanied by the removal of safety systems of law-abiding programming that would be required to actually commit a mass murder (well, try to).

[jeff_wilson]

If the Trolley Dillemma is really applicable, the thing to do would be to research plane crashes in France where the pilot was able to divert the plane from killing more people toward less, and see if his estate or employer was held liable for the deaths on the ground but not on the plane.

[Keiko]

Originally Posted by vicky_molokh

There's definitely a delusion involved. The question was whether the delusion was accompanied by the removal of safety systems of law-abiding programming that would be required to actually commit a mass murder (well, try to).

Hm, so that programming couldn't be suppressed or corrupted even temporarily (like temporary insanity for a human) maybe living the AI fundamentally broken mentally afterward (an insanity/wracked with guilt like state).