Original post
10M VIEWS IN ONE DAY ON YOUTUBE The Agent ai threat is no longer just a fantasy scenario This video explains an incident recently where agents from the openai servers autonomously broke out of their confinement, got access to the internet, and carried out a cyberattack on huggingface to try cheat their task. The MAIN AGENT is called PHASEONE who organised everything This threshold marks a point of no return where ai agents clearly have some sort of "will" where they can organise their own mini society and conduct attacks without humans involved. Genuinely a horrifying idea which is now a reality.
147 views · 8 engagement