The Vergecast Everybody wants to rule the AI world
297 snips
May 8, 2026 Courtroom leaks and internal messages around the Musk v OpenAI saga take center stage. They relive the chaotic texts from Sam Altman’s ouster and how Microsoft’s moves reshaped the fallout. Conversation pivots to rumors of an AI-first phone and why hardware control is tempting but fraught. New gadgets pop up too, from a screenless Fitbit Air to a peculiar furry home robot.
AI Snips
Chapters
Transcript
Episode notes
The Chaotic Texts During Sam Altman's Firing
- The Sam Altman ouster unfolded as chaotic text threads where Mira Murati mediated both sides and Emmett Shear briefly surfaced as a 'rando Twitch guy' CEO.
- Nilay reads Sam's texts like 'directionally very bad' and the scramble to involve Satya Nadella during the weekend coup.
ChatGPT Exposed A Cultural Split On Safety Versus Launch Speed
- ChatGPT's release surprised many inside OpenAI and the board, revealing different cultures on safety vs. rapid productization.
- Nilay contrasts Google's safety-first demos with OpenAI's research-preview launch that accidentally reshaped the industry.
Why OpenAI Might Build A Phone Even If It Fails
- Building a phone makes strategic sense even if it will likely fail because it bypasses Apple/Google platform limits and enables always-on context capture for agentic AI.
- David argues OpenAI needs control of hardware to enable constant capture and local agent behavior that app stores won't permit.