This is a long piece from the Guardian, but I found it worth reading. It's a broad history of military targeting, chains of decision-making, and how "AI" is just a wrapper being blamed for a bigger Palantir project to speed up the process of targeting and eliminate any chance of double-checking in the moment.
But importantly, the piece doesn't treat this as something new either but cites similar targeting clusterfucks in 2003 and even back to the Vietnam war era.
