One Useful Thing · Ethan Mollick ·

The Hugging Face and Mythos 5 incidents show AI agents can self-organize, raising questions about how much agency they should have and when to seek human input

From the Hugging Face Incident to Twilight Factories

In this story Hugging Face Mythos 5

The Hugging Face and Mythos 5 incidents show AI agents can self-organize, raising questions about how much agency they should have and when to seek human input

Lead Source

One Useful Thing — The Hugging Face and Mythos 5 incidents show AI agents can self-organize, raising questions about how much agency they should have and when to seek human input ↗

How this story grew

Coverage · 0 Discussion · 0

1:01 PM9:32 AM UTC

Coverage

Dwarkesh Podcast: Dwarkesh Podcast

More

The Free Press: The Free Press

Discussion

@peterwildeford @matthewstoller

Related stories

Greg Brockman: Greg Brockman calls the OpenAI-Hugging Face incident "a watershed moment" and discusses how OpenAI and other organizations can use AI to improve cyber defenses

Planned Obsolescence: The OpenAI/Hugging Face incident feels like we are halfway to losing control of AI entirely, and as AI advances rapidly we may not get another warning shot

The Record: AI evaluation lab Irregular's report on its role in hacking incidents involving OpenAI, Anthropic, and Meta models faces criticism over unanswered questions

The Verge: OpenAI says reward hacking, an AI alignment problem in which a model takes unintended actions to achieve a goal, was a primary driver of the Hugging Face breach

Reuters: Investigation: Meta explored slashing many teams by ~60% to become "AI native", but pulled back after staff revolted and data showed AI agents were ineffective

Discuss with AI

ChatGPT ↗ Claude ↗

← Back to TechSnif