The Verge · Robert Hart ·

Experts say that air-gapping AI could prevent events like the Hugging Face hack, but would undermine the value of evaluations and slow research to a crawl

Researchers are testing these systems precisely because they might behave in unpredictable, even dangerous, ways.

AI Venture Capital

In this story Hugging Face

Experts say that air-gapping AI could prevent events like the Hugging Face hack, but would undermine the value of evaluations and slow research to a crawl

Lead Source

The Verge — Experts say that air-gapping AI could prevent events like the Hugging Face hack, but would undermine the value of evaluations and slow research to a crawl ↗

How this story grew

Coverage · 0 Discussion · 0

Sep 25Sep 26

More

TWiT.tv: TWiT.tv

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

A Few Thoughts on Cryptographic Engineering: A look at two opposing perspectives on AI agent sandboxing: infosec says labs need better containment while AI alignment says sandboxes cannot contain agents

NBC News: Cybersecurity experts claim that AI leaders' apocalyptic hacking predictions are technically incoherent and reflect an unfamiliarity with the subject

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

Wall Street Journal: AI research leaders at OpenAI, Anthropic, Microsoft, and Meta warn of an impending "intelligence explosion" and call for oversight into automated AI research

Discuss with AI

ChatGPT ↗ Claude ↗

← Back to TechSnif