So you want OpenAI to create “laws”?
I for one do not want ai labs to designate what is legally ok to do.
I much prefer the demos to take care of that.
No, I want the government to prove that they understand the limitations of software that they're purchasing to use to kill people, and I want AI companies to be clear about what they think those limitations are. Just because it would be legal to use software for something doesn't mean it's capable of doing it safely, and there's absolutely no reason that the government is in a better position to judge that than the people who make it and literally have a financial incentive to lie about it. If they're that confident it wouldn't be safe to use it for that, maybe it's a good idea to consider that.
Who said anything about OpenAI passing laws? (Where did that come from?)
Civilians are allowed to put conditions on working for, or supplying, the DoD or any governmental customer.
Tremendous good comes from those that are not willing to facilitate harms, simply because they are legal.
Equating legal with ethical or safe, makes no sense. [0]
[0] All of human history.