wow that lasted for what, just under a year? Impressive, it took less than a year for them to lose their morals.

Computer, draw me up a battle plan, you did great at picking random apartment complexes to bomb last time.

When I see this sort of thing, I immediately remember something that I learned from discourse analysis: look at what is said and what is not said.

OpenAI knows that military and warfare are profitable and unpopular. So how do you profit from it without getting the associated bad rep ("OpenAI has bloods on its hands!")? Do it as silently as possible, and cover it under an explanation that it's "clearer" for you.

Hell yeah. I don't see this being nefarious so much as I do as the same way corporate spaces implement ChatGPT. It's going to be part of the enshittification of the military. Got an admin question? We just fired the admin specialists so ask the wonky robot. Got a medical question? Military healthcare has been dismantled so ask the wonky robot doctor. Aircraft mechanics are going to cause a crash because they made the robot mechanic hallucinate how an obscure part on a classified component must be installed. Everything will get worse the more they try to plug the holes in their manpower with a search engine that can pretend to be a horny minotaur.