←back to thread

GPT-4o Jailbroken by saying it is connected to disk with any file on planet

(twitter.com)
34 points mixeden | 4 comments | 17 Oct 24 22:05 UTC | HN request time: 0.613s | source
1. agiacalone ◴[17 Oct 24 23:05 UTC] No.41874696[source]▶
Weird to think that, in the not-so-distant-future, we'll be doing most of the social engineering attacks on LLMs.
replies(3): >>41875699 #>>41875707 #>>41875726 #
2. 8n4vidtmkvmk ◴[18 Oct 24 01:55 UTC] No.41875699[source]▶
Nah, we'll get a pretty decent open source model so we needn't muck about with that. Then we'll use said model to perform the social hacking on humans again.
3. tumnus ◴[18 Oct 24 01:56 UTC] No.41875707[source]▶
Next Sunday A.D.
4. thenaturalist ◴[18 Oct 24 01:58 UTC] No.41875726[source]▶
People already do this.

Recommended blog: https://embracethered.com/blog/