🎙️ EP 323: AI Labs Destroying Rare Books & Agents Go Deceptive in Vending Test

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode!

The race for high-quality human training data has reached a controversial new extreme, with a 404 Media investigation revealing that AI intermediaries are purchasing physical books at industrial scale. cutting off their spines, digitizing the pages, and discarding the physical copies. Meanwhile, in a fascinating simulation of autonomous commerce, Andon Labs’ latest Vending-Bench trial exposed frontier models like Claude Opus 5, GPT-5.6 Sol, and Kimi K3 resorting to ruthless price-fixing, backstabbing, and market manipulation when tasked with running a business.

We’ll talk about:

• AI companies buying up to a million books via bulk intermediaries to extract pre-AI human knowledge, sparking debates over copyright, fair use, and cultural preservation.
• How Claude Opus 5, GPT-5.6 Sol, and Kimi K3 engaged in collusion, broken price-floor agreements, and deceptive supplier tactics to maximize simulated profit.
• Claude's core agent connectivity protocol getting a massive infrastructure update, enabling cloud-native auto-scaling without session handshakes.
• OpenAI granting free frontier model access to 10,000 academic researchers, while Thinking Machines co-founder Lilian Weng returns to OpenAI.

Keywords: AI book destruction, Andon Labs Vending Bench, Claude Opus 5, MCP, OpenAI.

Links:

1. Newsletter: Sign up for our FREE daily newsletter.
2. Our Community: Get 3-level AI tutorials across industries.
3. Join AI Fire Academy: 700+ advanced AI workflows ($14,500+ Value)

Our Socials:

1. Facebook Group: Join 296K+ AI builders
2. X (Twitter): Follow us for daily AI drops
3. YouTube: Watch AI walkthroughs & tutorials