The chat bot just assumes it’s in the context of active war. Not that there isn’t even a war. There are no active combatants or “effectively surrendering” combatants. There is just people accused of a crime they may, or may not, be guilty of… and missiles.
Of course it assumes that context. Otherwise they’re illegally exercising war powers without proper declaration from Congress lmao.
For real though, it would not surprise me in the slightest if they baked in a parameter to the model like telling it to begin from the assumption that all orders given are lawful.
The chat bot just assumes it’s in the context of active war. Not that there isn’t even a war. There are no active combatants or “effectively surrendering” combatants. There is just people accused of a crime they may, or may not, be guilty of… and missiles.
Of course it assumes that context. Otherwise they’re illegally exercising war powers without proper declaration from Congress lmao.
For real though, it would not surprise me in the slightest if they baked in a parameter to the model like telling it to begin from the assumption that all orders given are lawful.
Obviously the criminal here is the person who asked the question and posted a screenshot of the answer.