#172 - Claude and Gemini updates, Gemma 2, GPT-4 Critic
Summary
This episode covers Anthropic’s new Claude features, Google’s Gemini integration, and OpenAI’s delayed voice mode. It also discusses MIT’s specialized AI chip, Huawei’s chipmaking investments, and ByteDance’s partnership with Broadcom. Finally, the hosts address Chinese firms migrating from OpenAI APIs and OpenAI’s revised stock sale policies.
Topics discussed
Sponsor message and episode intro
Host introductions and show housekeeping
Anthropic updates and UX developments
Google Gemini integration across Workspace
OpenAI voice mode delay and Apple collaboration
Waymo expansion and Figma AI features
Sohu ASIC chips and specialized hardware
China's semiconductor supply chain race
Chinese AI models and ChatGPT alternatives
OpenAI secondary share controversy
AI for compiler optimization research
Google Gemma 2 and ESM3 protein model
OpenAI CriticGPT and code verification
ChatGLM benchmarks and scaling laws
DeepMind's DNA editing breakthrough
Chinchilla scaling laws and safety alignment
Senate Bill 1047 and AI regulation
Rwandan AI disinformation campaign
Jailbreak attacks on LLM safety filters
Music industry lawsuit against AI training
Toys R Us Sora ad and episode outro
Listen ad-free on Castria