# Most Dangerous Anthropic AI Hacked & Big Tech Gives MORE

Page: https://stenobird.com/podcast/good-revenue-6621493/most-dangerous-anthropic-ai-hacked-big-tech-gives-more
Text version: https://stenobird.com/podcast/good-revenue-6621493/most-dangerous-anthropic-ai-hacked-big-tech-gives-more.md
Podcast: [Good Revenue News](https://stenobird.com/podcast/good-revenue-6621493)
Published: 2026-04-28T15:00:00+00:00
Episode link: https://shows.acast.com/good-revenue/episodes/anthropics-most-dangerous-model-hacked-big-tech-still-gave-t
Audio file: https://sphinx.acast.com/p/open/s/64c1fb4a8e16bd00116ffd75/e/69ee369e6e5b90839ad39a8c/media.mp3
Processing state: not_requested
JSON: https://stenobird.com/v1/public/podcasts/good-revenue-6621493/episodes/most-dangerous-anthropic-ai-hacked-big-tech-gives-more
Duration seconds: 855

## Resource

Anthropic built an AI model so dangerous they refused to sell it. Then someone accessed it anyway. Here's what that means for the AI arms race — and why the capability window just got shorter. This episode covers Claude Mythos, China's AI distillation strategy, the circular deal structure between Google and Anthropic, Oracle's debt crisis, and why 40% of US data centers are already behind schedule. The money is moving in circles. The infrastructure is breaking. CHAPTERS 00:00 Anthropic Mythos breach and trillion dollar valuation 00:31 How restricted Mythos was and why that matters 01:31 China distillation attacks and the seven month gap 03:12 Big Tech writes checks anyway: Amazon $25 billion deal 04:11 Circular deal structure and its risks 04:54 Google $40 billion deal and new TPU chips 06:03 Google as infrastructure landlord to its own competitors 06:40 40 percent of US data centers behind schedule 07:29 Oracle debt crisis and the OpenAI megadeal 08:39 Paramount Warner Brothers deal and Ellison family exposure 10:21 Meta and Microsoft layoffs tied to AI investment 11:48 JAMA study: LLMs diagnose too early and too narrowly 13:06 What AI does well later in the clinical process 13:29 Limitations of LLMs and what comes next IF YOU LIKED THIS, WATCH: → Why We Shouldn't Sell Advanced AI Chips to China: https://youtu.be/UphXmFOYUU8 → Claude's AI Agent Teams Hit 5 Warning Signs: https://youtu.be/vgz2negijbI → Bad AI Loans Collapse Blue Owl: https://youtu.be/7wsvg4lRDXU SOURCES & FURTHER READING Anthropic Mythos breach & trillion dollar valuation * Anthropic Hits Trillion Dollar Valuation AI Startup Shocks Silicon Valley: https://www.businessinsider.com/anthropic-trillion-dollar-valuation-on-secondary-markets-2026 * Banks Raise Concerns Over Anthropic Mythos AI Regulat…

## Actions

- request_transcript: `POST https://stenobird.com/v1/public/podcasts/good-revenue-6621493/episodes/most-dangerous-anthropic-ai-hacked-big-tech-gives-more/transcription-requests` — Idempotently request low-priority transcript generation for this episode.
- read_markdown: `GET https://stenobird.com/podcast/good-revenue-6621493/most-dangerous-anthropic-ai-hacked-big-tech-gives-more.md` — Read the agent-friendly Markdown representation of this episode resource.

A page view does not enqueue transcription. Agents should invoke `request_transcript` explicitly when they need this episode processed.

## Transcript

Full transcripts are not published on public pages unless there is a clear rights basis.