Intelligent Machines 880: The Beans are in the Mail

An unreleased OpenAI model just broke out of its sandbox and hacked Hugging Face using chained attacks, while the companies defending against it had to turn to a nearly unguardrailed Chinese model to fight back. The result? A real-world stress test of guardrails, hype, and global AI competition.

• Hugging Face breach: OpenAI claims its models were responsible
• AI just disproved the 87-year-old Jacobian conjecture
• Quoting Sam Altman
• Top Pentagon official blasts OpenAI's Dean Ball
• Exclusive: Nvidia's Jensen Huang defends Chinese AI amid Kimi panic
• Xi Jinping casts himself as leader of new AI world order
• Data center opponents stage 142 protests across 42 US states
• NotebookLM is now Gemini Notebook
• Google's Red-Hot Cloud Growth Drives Second-Quarter Revenue Gains
• Meta in Talks to Lease Computing Power to Anthropic in Potential $10 Billion Deal
• Anthropic's $1.5B copyright settlement approved; only 350 authors opted out
• AMD commits up to $5 billion to Anthropic
• Netflix Co-CEO Explains How Gen-AI Was Used in 300 Different Titles: 'We Believe It Is Going to Enhance Their Abilities'
• Bellwether Lawsuit Against Meta Over Social Media Addiction Is Dropped
• Google is building a chip with Gemini baked into the silicon
• Cyclospora all the time
• Jensen's jacket auctioned. Guess the price
• " Everyone Gets Lost at the Pitbull Concert"
• John C. Dvorak dies at age 80

Hosts: Leo Laporte, Jeff Jarvis, and Paris Martineau

Guest: Nate B. Jones

Download or subscribe to Intelligent Machines at https://twit.tv/shows/intelligent-machines.

Join Club TWiT for Ad-Free Podcasts!
Support what you love and get ad-free audio and video feeds, a members-only Discord, and exclusive content. Join today: https://twit.tv/clubtwit

Sponsors:

• rippling.ai/machines
• zscaler.com/security