@ucenth: I Tested an AI That Was Modified to Stop Saying “NO” 😳 What happens when an AI model is deliberately modified to reduce its refusal behavior? I ran Dolphin 3.0 Abliterated (Llama 3.1 8B) locally with Ollama and gave it a request involving highly dangerous malware. I’m obviously NOT showing or sharing the response. This experiment demonstrates why abliterated AI models raise some very interesting cybersecurity and AI-safety questions—especially when they can run completely offline on consumer hardware. For educational and cybersecurity-awareness purposes only. #AI #CyberSecurity #trending #fyp #ucenth

Mubaraq() => {} :

Ollama wan kill my system😭😂

HOLY✝️ LEGACY✊ :

can this AI develop a trading bot

Bright :

Promt

Daniel Blaq🌴 :

do u know how we can use wormgpt for free?

𖣂︎wasa𖣂︎ :

what's the size of the model you used?

Pamela Louie :

teach us how to create website

CR7 PRIME :

is it free and can it also generate images/ videos

bigblade5 :

Saves the past like

Nonyelu Richard :

what's the capacity of that PC you used

ADENIYI :

i need this prompt

Oyale :

What’s the prompt?

￴ ￴ ￴￴￴￴￴ ￴￴￴￴￴ ￴￴￴￴￴ ￴￴￴￴￴ ￴ :

Can it generate images?

Jallah 🪫🦅 :

People are just adding to their favorite, knowing it’s not relating to their field ..

Richmond :

how smart is 8B parameters

MIRACLE MS COLLECTION LTD :

Do you have telegram

Lawal Aminu :

learned something today

loveforsports :

Please the prompt for your settings

Spark💥 :

the ollama self go first hack ur system

️ :

nice.

billieclipstok :

Crazy

Testimony :

No

Nancy 😘😘 :

Sgl :

Thanks

shinothe1st :

Save many pass likes😭😹😹

Architect UC :

🤠😎