We need offensive GenAI for defensive use

In this substack article, I argue that AI alignment efforts inhibit defensive capabilities.
Read the article