AI risk, policy, and live show plans

Brought to you by TogetherLetters & Edgewise!

In this episode:

AI Models & Big Money

• Anthropic Is Testing 'Mythos,' Its Most Powerful AI Model Yet
• Anthropic Leaked 500,000 Lines of Its Own Source Code
• OpenAI Raises Record $122B at $852B Valuation
• Google Employees Have a New AI Tool Called 'Agent Smith'
• Meta Is Running Intensive AI Training Weeks for Employees

AI Policy & National Security

• Trump Administration Unveils National AI Policy Framework
• Super Micro Computer Co-Founder Charged With Diverting $2.5B in Nvidia AI Chips to China
• Philadelphia Courts Will Ban All Smart Eyeglasses
• AI and Bots Have Officially Taken Over the Internet

Tech & Innovation

• Fannie Mae to Accept Crypto-Backed Mortgages for the First Time
• Order Sunlight Whenever and Wherever You Want on Demand
• Humanoid Robotics Maker Sunday Reaches $1.15B Valuation

Weird and Wacky

• A Man Used AI to Call 3,000 Irish Bartenders to Track the Cost of Guinness
• YouTuber Resolves Fraternal Disagreement With Excavator and Hilariously Large Sword
• High-Tech Olaf Animatronic Collapses Mid-Performance at Disneyland Paris
• April Fool's 2026: The Most Ridiculous Tech Pranks of the Year

Tech Rec:

• Sanjay - Parachord
• Adam - CoPilot budgeting app