Re:GPT-4 Details Leaked
by namaria on Tuesday July 11, 2023 @12:38PM
(#36683489)
"I can't conceive of such failure mode thus we're safe"
[ Parent
| on HN ]
Sorry, I do not understand what you mean...
I think they read the first line of your reply and took it as you arguing that it's nearly impossible for an LLM to give output that would get someone killed.
I was agreeing with them. Why assume wilful ignorance? Have we become the new reddit? People just yelling in disagreement? Maybe it's time to log out and delete this password...
No, Namaria, we were just trying to understand what you meant - which for us resulted cryptic.