2025-11-02
Satya on BG2: “Absolutely Azure could have grown ~41-42% if we'd had more compute” 👀 h/t @TMTBreakout for the notes
@bg2pod
Q&A with Sam Altman and Satya Nadella about the Microsoft-OpenAI partnership, OpenAI's restructuring and $100B revenue target for 2027, $3T AI buildout, more
BG2. All things AI w @sama & @satyanadella. A Halloween Special. 🎃🔥@bg2pod @altcap — (00:00) Intro (02:28) Microsoft's Investment in OpenAI (03:19) The Nonprofit Structure and Its ...
2025-01-27
Everyone has the DeepSeek thing literally ass-backwards. The hyperscalers/frontier LLMs will learn/use whatever they can from it (along w everything else already on their roadmaps) to make their models even better and then exponentiate them with these looming giga clusters.
YouTubeTranscriptOptimizer
A bear case for Nvidia: hardware competitors, LLM code translation to avoid CUDA lock-in, DeepSeek's training and inference efficiency breakthroughs, and more
DeepSeek recently released models claiming up to 45x more efficient training and inference compared to today's best-known large language models (like OpenAI's o1). … Forums: Hacker...