Video by Better Stack on August 28, 2026. May be an image of text that says 'README AGPL-3.0liconse Heretic: Fully automatic censorship removal for language models Lar 9 193 ONLINE MATREX CODEBERGHIRROR CODEBERG MIRROR Follow OTHLETRENDING TRENDING Repository The Day THIS TOOL IS DANGEROUS! multim models, DIcOd pace even some hybrid modelsike andcertain other research supported uto the bor. BOGBPCI 80GB PCIe pt-015-206 206 er'.
betterstackhq's profile picture
betterstackhq
Verified
•
Follow
More options
betterstackhq's profile picture
betterstackhq
Verified
 
This Open Source Tool Removes AI's Safety Filters
im.hidden667's profile picture
Reply
maxwvm's profile picture
maxwvm
 
I mean there are open weight models without safeguards; safeguards can only be added post training, think of this like lobotomising the system. Open weight models will just jo this.
Reply
sariktillko's profile picture
sariktillko
 
What if tools required licenses that are tethered to real verified human?
Reply
jacks.ej255's profile picture
jacks.ej255
 
AI slop, can even pronounce the library
Reply
useraccessdenied_'s profile picture
useraccessdenied_
 
A B Literation LoL AI
1 like
Reply
laugatjanaz's profile picture
laugatjanaz
 
It’s called npm. It’s the most dangerous
3 likes
Reply
Like
166
Comment
7
Share
Save
Log in to like or comment.
More posts from betterstackhq
See more posts