Dario and Anthropic's strategy has been to exaggerate the harmful capabilities of LLMs and systems driven by LLMs, positioning Anthropic themselves as the "safest" option. Take from this what you will.

As an ordinary human with no investment in the game, I would not expect LLMs to magically work around the well-known physical phenomena that make submarines hard to track. I think there could be some ability to augment cybersecurity skill just through improved pattern-matching and search, hence real teams using it at Google and the like, but I don't think this translates well to attacks on real-world targets such as satellites or launch facilities. Maybe if someone hooked up Claude to a Ralph Wiggum loop and dumped cash into a prompt to try and "fire ze missiles", and it actually worked or got farther than the existing state-sponsored black-hat groups at doing the same thing to existing infrastructure, then I could be convinced otherwise.