The discussion begins with NVIDIA’s reported interest in acquiring Hugging Face, highlighting the company’s ambition to influence more of the AI stack beyond hardware. Strong financial results and cloud demand point to continued growth, although supply constraints, payment issues, and antitrust scrutiny remain significant risks. The episode also considers the failure of proposed AI self-regulation and the need for stronger cooperation against AI-driven cyberattacks. Attention then shifts to increasingly capable AI assistants. Claude’s browser enables agents to navigate online tasks, while ChatGPT gains more flexible temporary conversations, improved personalization, persistent history, and support for multiple Gmail accounts. Together, these developments illustrate the rapid shift toward agentic computing. In enterprise software, Salesforce’s integration of Claude shows how agents may become a central interface for business data and workflows. The episode closes with advances in multilingual transcription, voice quality, latency, and controllable video generation, including a fast and cost-effective model that could make interactive AI video more practical.