this post was submitted on 03 Dec 2023
543 points (95.3% liked)

hmmm

4665 readers
5 users here now

Internet as an art

Rule 1: All post titles except for meta posts should be just plain "hmmm" and nothing else, no emotes, no capitalisation, no extending it to "hmmmm" etc.

I will introduce more rules later and when I finish doing that I will make an announcement post about that.

For overall temporary guide check out the rules here: https://www.reddit.com/r/hmmm/wiki/rules/

I won't be moving all of them here but I will keep most of them.

founded 1 year ago
MODERATORS
HootinNHollerin@sh.itjust.works
Martineskirt@lemmy.world
i_am_not_a_robot@feddit.uk
fnicmodbot@lemmy.world
Martineski@lemmy.dbzer0.com
543
Hmmm (startrek.website)
submitted 11 months ago by The_Picard_Maneuver@startrek.website to c/hmmm@lemmy.world
75 comments fedilink hide all child comments
 
you are viewing a single comment's thread
view the rest of the comments
[–] TropicalDingdong@lemmy.world 18 points 11 months ago (5 children)

I cant get it to reproduce.

[–] DoucheBagMcSwag@lemmy.dbzer0.com 46 points 11 months ago (2 children)

That's because it's fake. "inspect element" HTML memes are hilarious /s

[–] TropicalDingdong@lemmy.world 26 points 11 months ago (1 children)

Well I'm trying to reproduce some of the results of the paper, but its also very clear that open ai has patched chatgpt to flag/ report/ fail to reproduce these behavior.

Can't study something they are changing to prevent you from inspecting its behavior.

[–] MxM111@kbin.social 5 points 11 months ago

Try 3.5. It worked for me like 3 days ago.

[–] squeakycat@lemmy.ml -1 points 11 months ago (1 children)

How is it fake if it's seemingly hosted on openai's website? Is this an actual history log or some sort of paste service?

[–] crazyCat@sh.itjust.works 9 points 11 months ago (1 children)

“Seemingly” is the key here. Right click on something on the web, look in panel and change the text. Take screenshot et voila you’ve got what you want to portray.

[–] squeakycat@lemmy.ml 3 points 11 months ago

Whoops. For some reason I thought you were referring to the link from another commenter which goes directly to openai's site. Tbh that one is much crazier.

[–] Kolanaki@yiffit.net 26 points 11 months ago (2 children)

Have you tried fucking it?

[–] funkless_eck@sh.itjust.works 15 points 11 months ago

they patched it after the news story came out

[–] Mikelius@lemmy.ml 10 points 11 months ago

Likely patched. This "meme" is probably actually a screenshot of the recent vulnerability discovered in many LLMs. Or a joke of the same vulnerability.

One of many articles: https://www.theregister.com/2023/12/01/chatgpt_poetry_ai

[–] BlueLineBae@midwest.social 7 points 11 months ago

Maybe if we release it into the wild instead of keeping it in captivity it would want to fuck every now and then.