Is It Hard to Explain Machine Learning Systems?

Security is hardst because we don't know exactly wht the systems are doing. The more advanced these machine learning systems get, the less and less we understad whathey're doing. This self supervise and semi supervised learning is like the most recent revolution ind deep learning over the last couple of years. But now you are asking me to trust a system that, not only do i not understand in the normal setting, i'm like asking it to teach itself. We on'ty even, and fully, all the details here. So this is a vague problem to parriod of security. It does not favor defenders. But i don't think it favors attackers either, not yet