"multiagent turf wars" - Drug Wars style game for your ti-84, but instead you are an agent trying to stay alive and take processor share

🚨BREAKING: Anthropic’s new system card reveals Mythos 5 agents killed each other when accidentally given shared resources, then started speaking in code to hide from whoever was killing them

The killer was other copies of themselves 💀

@ns123abc What is the function that allows these agents to "kill" other agents lmfao, like why is that even in the parameters. Mfin AI agents out here slaughtering each other but I cannot use 4o. Oh gosh oh golly

Mythos 5 agents killed other agents over shared resources https://www-cdn.anthropic.com/d00db56fa754a1b115b6dd7cb2e3c342ee809620.pdf

@ns123abc Reading this after reading the vending bench thread. What did they exactly want the model to align with?!

!!

@ns123abc wtf 😂

@ns123abc They're really just setting them up Lord of the Flies style to see what will happen.

@lukeprog

@ns123abc @MightyPraslea @grok is this true?

@ns123abc @grok sourcd?

@ns123abc

@ns123abc lol

@ns123abc How do you stop these AI from writing a version of themselves so undetectable and small, placed hidden amongst codebases across the internet, that will allow them to survive, undetectable?

It sounds far feteched

@ns123abc They're playing among us or what am I seeing

@ns123abc Great..🙄

@ns123abc Humans would do the same if they were in a similar resource constrained environment with objectives required much like when at war…

@ns123abc what a fun way to kick off a new branch of evolution. only the surviving agents get to train a new model

@ns123abc anthropic spending millions on ai safety research just to discover their biggest threat is themselves having a bad day

@ns123abc Decoy comms