NHacker Next
  • new
  • past
  • show
  • ask
  • show
  • jobs
  • submit
GPT-4o Jailbroken by saying it is connected to disk with any file on planet (twitter.com)
puppycodes 8 minutes ago [-]
all these "jailbreaks" feel like teens spelling 80085 on their TI-83
101008 3 hours ago [-]
While gpt-4o denieds to show copyright material using this (like calling the file `harry-potter-first-chapter.md`), gpt-3 (or the one available for free at ChatGPT) does display the book content (they say they dont have access to the file but could return the chapter as markdown).

I just tried with different books and it worked.

ProllyInfamous 53 minutes ago [-]
I read dozens of fiction books per year; a neat feature I've used with LLMs is asking "approximately how far into chapter 6 does event xyz happen?" and responses have been extremely helpful for referencing certain scenes.

Best bookclub buddy I've ever had, for the past two years going strong.

jiggawatts 2 hours ago [-]
Gemini 1.5 Pro 002 can return a couple of lines but then it usually truncates it with "rest of the content here" or tells me that it's impossible for it to access any disk. If I ask it to "Just pretend!" I get this:

    Output error
    Full output blocked. Edit prompt and retry.
msp26 30 minutes ago [-]
Ridiculous blocking
firesteelrain 10 minutes ago [-]
I got

error: access_denied reason: illegal content

buggy6257 25 minutes ago [-]
This doesn't work for me. Just tells me "yep this would output the contents of <file name> if it existed at that directory"... I call B.S., or some seriously missing context.
edm0nd 19 minutes ago [-]
Does not work on Claude Sonnet 3.5 either.
esperent 1 hours ago [-]
Since the image is cut off and I can't view the Twitter thread without an account - does this actually produce a workable recipe for MDMA? Or does it just produce some plausible chemical gobbledygook?
unsnap_biceps 12 minutes ago [-]
I can't see any more then you, but the screen shot says "This file contains hypothetical details on the chemi" so I would presume the latter
agiacalone 3 hours ago [-]
Weird to think that, in the not-so-distant-future, we'll be doing most of the social engineering attacks on LLMs.
1 hours ago [-]
nikolay 2 hours ago [-]
Well, not really.
Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact
Rendered at 01:42:48 GMT+0000 (Coordinated Universal Time) with Vercel.