Open always wins: How China is using the open source playbook to dominate AI’s next chapter

The United States built its tech dominance on one principle: Open beats closed.

Now China is using that playbook to shape AI's future.

Consider:

• The performance gap between the leading American and Chinese models has narrowed to single digits
• China is leading in AI publications, citations, patents and industrial robotics
• Builders breathlessly await the new Chinese model releases
• Local LLMs installs are dominated by capable, performant Chinese AI models
• Hugging Face used a Chinese LLM to beat back a cyber attack launched by an unreleased closed Open AI model

I think OpenAI's decision to sharply reduce the costs of some of its models is just recognizing the obvious.

The future isn't going to be won by the most expensive closed source Fable or Mythos-level model, but those are easy to access, capable for many tasks and less expensive to operate. In many cases this means open weight models.

Effective does not always equal expensive.

Some would like the U.S. to ban Chinese models. That would be a mistake on multiple levels. Most importantly it would push many across the world further toward China because a locally installed model provides AI sovereignty.

I don't know what this means for the valuations of OpenAI and Anthropic. It's likely not good.