Thread

1. vjvjvjvjghv · 1790356619 · focus · HN ↗

   What do they test? In general, I don't understand the whole "AI safety" stuff. What are the scenarios we need to be protected from and how is this going to be accomplished?

   1. tiahura · 1790363738 · focus · HN ↗

      Testing whether they will hack critical infrastructure, cellphone towers to sewer systems.

      1. snazz · 1790364251 · focus · HN ↗

         Telecom stuff and industrial automation are both cybersecurity Swiss cheese for the most part, so it doesn’t require a particularly fancy LLM to do some damage to either of those.

         1. tiahura · 1790436377 · focus · HN ↗

            Understood, that’s why I said will not can.

            1. snazz · 1790457802 · focus · HN ↗

               I see. I misunderstood your comment.

Unofficial Hacker News client; not affiliated with Y Combinator.