> But the current structure of defense funding makes this nearly impossible. VC-backed defense startups aren't the answer either. They're making the same mistakes - small compute, off-the-shelf models, requiring relocation from experienced 40+ year old scientists who won't move. They're essentially just spending the money the government can't, without solving the fundamental issues.
Have you tried to express your perplexities to one of the DARPA PMs?
Theoretically, a significant part of their work should precisely consist of receiving feedback on the mistakes made in order to iterate more quickly toward effective solutions.
Have you tried to express your perplexities to one of the DARPA PMs?
Theoretically, a significant part of their work should precisely consist of receiving feedback on the mistakes made in order to iterate more quickly toward effective solutions.