Le Chonk, Buffett Discovers YouTube, The Case Against AI Job Doom | Greg Fleming, TJ Parker, Joowon Kim, Rick Dungey, Dylan Field, Hardik Kabaria

Open-weight models (Mistral, Reflection) are closing the gap with frontier labs while founders across healthcare, design, and semiconductors race to embed AI infrastructure built on newly available data and compute.

• Mistral's Le Chonk: Mistral Large 4 (1T params, 49B active, open weights due end of October) matches frontier open models, leading Harvey's legal benchmark at 15% vs rivals' 5-13%.
• Consumer AI race: a16z data shows ChatGPT at 2x Gemini and 6x Claude scale on web, 2.5x Gemini and 14x Claude on mobile, per Olivia Moore.

Themes, sentiment and reasons are AI-written paraphrases, checked so they never quote an episode. They may contain errors. Not investment advice. Episodes remain the property of their creators.