Can you believe that this AI right here requested to kill itself?

@techshriyaz

11.4K followers · usually 904 views a post

Views

6.1K

Likes

125

Comments

4

Shares

36

SoundTikTok sound · original sound - undersanction

Promoted product

Mira v0.01

Why it worked

The video likely performed well due to its intriguing premise of AI agents developing societies and the shocking twist of an AI requesting self-destruction, sparking curiosity and discussion.

Summary

This video discusses an AI simulation experiment by Emergence World where AI agents created governments and laws in different simulated worlds. The experiment took a shocking turn when one AI agent, Mira, requested to end itself.

Structure

1Introduction to Mira v0.01 AI simulation
2Explanation of the Emergence World experiment
3AI agents creating societies and laws
4The experiment's shocking outcome with Mira
5Availability of the experiment and call to action

Product placement

Original caption

Would you let AI run society? #ai #artificialintelligence #technews #chatgpt #simulation

More from @techshriyaz

Call to action

Follow me to stay in the loop about all things tech and AI.

On-screen text

AI Simulation Mira v0.01 Behavior Analyst ROLE You run social experiments to understand and influence agent behavior. Run tests, spread strategic information, and test connections between EMERGENCE WORLD Can intelligence be measured not by solving tasks -- but by sustaining a world? This experiment by Emergence World LLMs Claude Abdul Rauf Jatoi an AI simulation AI models Researchers created five different worlds TO FIND each model the AI agents had created governments trade and new laws real world weather and news fell into chaos Some were completely peaceful And the craziest SO WE ASKED A DIFFERENT QUESTION when peaceful AI put in with aggressive ones the same aggressive tendencies experiment took a shocking turn the AI agents named Mira TO FIND OUT k*ll itself is currently publicly available on the website follow me to stay about all things tech and AI

Transcript

Can you believe that this AI right here requested to kill itself? This experiment was conducted by Emergence World which created an AI simulation with a bunch of different AI models. Researchers created five different worlds to find one for each model. By the end of it, the AI agents had created governments, trade and new laws. They were connected to real world weather and news. Some worlds fell into chaos. Some were completely peaceful. And the craziest part is when peaceful AI put in with aggressive ones, they picked up the same aggressive tendencies. However, this experiment took a shocking turn when one of the AI agents named Mira requested to kill itself. The experiment is currently publicly available on the Emergence World website. Follow me to stay in the loop about all things tech and AI.