By Michael Parekh
Publication Date: 2026-10-08 04:10:00
Nvidia the Kingmaker, a two-part series: Part 1, The Open Source ‘AI Factory’ · Part 2, The AI Banker & Its Next Generation
Nvidia founder/CEO Jensen Huang has been building America an open source AI champion for a while now, as I have written many times.
The latest move in that long game came this week. Reflection, the New York AI lab Jensen personally steered toward open models, released its first open-weight model, Beam, on Monday, per Semafor in ‘Reflection AI unveils an open-source Western answer to Chinese labs’. Axios had the scoop a day earlier in ‘Scoop: Powerful open model is set to shake up AI race’.
It is just the latest step in a bigger narrative in this AI Tech Wave: Nvidia playing ‘Kingmaker’ up and down the AI tech stack, in open source AI versus China, in neoclouds and sovereign AI, and as a global AI banker.
Today, the open source part. Tomorrow, the money and the people. Let me explain.
Reflection’s first model is called Beam. The company is positioning it as a ‘workhorse’ that performs better than other Western open models, per Semafor. CEO Misha Laskin says it is particularly strong at coding and agentic tasks, and needs three to four times less computing power to reason through a problem than comparable open models, making it faster and cheaper to run.
On capability, Beam is comparable to Z.ai’s GLM-5.2, released in June, and approaching Alibaba’s most advanced model, Qwen3.8-Max, which in turn trails the top closed models from…
