Our AI Cold War

Donald Trump has renamed artificial intelligence "super intelligence" (SI) because he thinks "artificial" makes it sound fake. His chief AI advisor is David Sacks, a venture capitalist and co-host of the All-In podcast, who served as Trump's White House AI and crypto czar for a little over a year and remains a close outside adviser. Sacks is pushing a vision for American AI that pits him against the country's two leading companies—OpenAI and Anthropic—and aligns him with China's open-source AI strategy. This inside-the-Beltway scramble is unfolding against a backdrop of genuine U.S.-China competition over the technology, with both countries racing ahead while safety concerns get pushed to the margins.

The open-source crusade inside the White House

David Sacks is no fan of OpenAI or Anthropic. He calls Anthropic "the porcupine" and warns it could become "the most powerful monopoly ever created in human history." His hostility stems from a fundamental disagreement about how AI should be built and distributed. The dominant model today is closed and proprietary: you sign up with a company like OpenAI or Anthropic, and you use the product they host on their servers. Sacks champions the opposite approach—open-weight or open-source AI—where the model's parameters and weights are published online for anyone to use, modify, or run themselves.