More than 200 human rights and civil society organizations are calling on governments and tech companies to stop the use of AI for military purposes, warning that it turns war into a "death conveyor" without sufficient legal frameworks and transparency.
They don’t need to ban AI, they need to enforce responsibility for “War Crimes.”
Doesn’t matter if you used AI to kill schoolgirls or not, the fact is: you were in command and you ordered a strike that killed schoolgirls. The transistor in the circuit board of the guidance computer is not liable, the human making the decision to go ahead with the strike is.
what about both?
Sue a transistor, see how it conducts itself.
false equivalence. that’s like saying in defense of war crimes, sue the weapons and war vehicles, and see how they defend themselves in court.
And how is that false equivalence to an algorithm running on a bunch of transistors?
because the point of my example is that neither case works like that. you can’t sue weapons, you sue those who commanded their use. you can’t sue transistors… and what does that mean? it means you sue those that commanded the use of AI assisted weapons.
common sense, really.
Reread the thread and point out anywhere you get even the slightest impression that I disagree with what you just wrote…
I read this as implying whwt I say does not make sense:
A - it’s a bad semiconductor pun
B - who in their right mind would think that a transistor is a legitimate target of a lawsuit?
C - all “AI” is is a pile of transistors with various charge states representing the software layers. It was built by people, same as a Smith & Wesson 38, nobody is suing the guns for people they’re killing, either, not even when they have fault sights and hit unintended targets.