Superman doesn't kill
We're talking about AIs here, but it feels like a very human problem as well.
If you think about the army going to fight a war in a particular place where there are rebels, the army guys are usually bound by certain laws. Because if not, you're going to be sent to the Hague for crimes against humanity. Yes, you can try and save yourself, but if you try and save yourself in a drastic way, you'll get put into court. But the rebels are not bound by those rules.
Even in superhero movies, you find that the good guy doesn't want to kill. Superman doesn't kill. But the person that is trying to do bad stuff doesn't have those rules, and so he's going to kill you. So how do you balance that out?
And we're seeing that play out in the AI world, which is hilarious.
Said out loud in a team meeting, 24 July 2026. Trimmed for filler; not rewritten.