Good morning. Here is what mattered in AI today. Nvidia’s board is reportedly preparing to remove chief executive Jensen Huang as a second allegation of misconduct surfaced, though the company has not announced any decision. Google launched a dedicated decision model that scores up to 1,000 questions in a single call, and TSMC’s board approved up to $1 billion in stakes in Anthropic and OpenAI. We also look at careers in AI evaluation and red-teaming, explain how vLLM splits multimodal serving across GPUs and profile the British inference-chip start-up Fractile after its latest funding round.