|
|
|
|
|
|
by simoncion
7 days ago
|
|
|
> Help me understand. Okay. Read these comment threads, then tell me if they change your understanding of what I've been talking about in this thread: [0][1] I'll address the rest of your comment after you get back to me. [0] <https://news.ycombinator.com/item?id=48999644> [1] <https://news.ycombinator.com/item?id=48999415> [2] [2] Yes, I'm aware that that one is the one we're talking in right now. You should re-read it with both the context from [0] in mind, as well as your initial comment to which [1] is a direct reply to, namely: > Why should OpenAI (or any frontier lab) be building these systems if they can't get a secure environment / containment right?
Because we continue to have zero evidence that aligment is an actual risk.
|
|
|