> Through the history of computer security, there has usually been a moat of expertise separating a vulnerability from its exploitation. ... Large Language Models (LLMs), the systems behind chatbots like ChatGPT, are building bridges across that moat.
Draining it, more like.