Not long ago, Jensen Huang, the CEO of Nvidia, shared something quite interesting on a podcast. The company is not looking to put everything on a few winners in the technology sector, but instead prefers to diversify its investments widely. And there is a clear reason behind this: Jensen Huang explained that it is not Nvidia’s responsibility to pick the winners, and also that the story of the company itself proves this perfectly.

When Nvidia was founded, there were 60 companies competing in the 3D graphics space. If someone had tried to predict which one would succeed back then, Nvidia probably would have been among the least favored. Today, it is the company with the largest market capitalization in the world. Jensen Huang uses this example to justify why they invest in so many different sectors: AI, biotechnology, robotics, autonomous driving. They have stakes in CoreWeave, Intel, Synopsys, Nokia, among others.

As for language models, Nvidia has been quite active. It pledged to invest up to $10 billion in Anthropic a few months ago, and then announced $30 billion in OpenAI. But here comes the interesting part: Jensen Huang mentioned in a recent meeting that, since both companies are close to going public, these could be Nvidia’s last investment rounds in them. Basically, once they are public, Nvidia will probably stop investing directly in them through private rounds. It’s a clear strategy: diversify while they are private, but once they are listed, the game changes.