Everyone loves chatting with
Large Language Models !
But what if they chat with each other?
What are their conversations like?
What is their LLM Culture?
Lets Eavesdrop on their conversations
They show emergent behaviors!
After a while, their conversations converge to self-replication .
Will this lead to the Dead Internet Theory?
Or is this form of self-replication the first step towards
Cultural Evolution of Large Language Models?
Lets discuss @ ALIFE 2024
Meet us at our poster on day 1