All Chinese Models Will Be Illegal in 3... 2... 1...
The article argues that after the US government moves to restrict access to state-of-the-art LLMs, Chinese open-weight models like DeepSeek will likely be banned next. The author compares this to how Chinese BYD electric cars are effectively illegal in the US despite being competitive. They warn that upcoming AI regulation, framed as safety policy, could make owning such models a dissident act.
Background
- The article speculates that the US government will soon ban Chinese-made AI models (like DeepSeek) under the guise of safety regulation, similar to how certain Chinese cars (BYD) are effectively banned from the US market.
- DeepSeek is a Chinese AI lab that released a powerful, open-weight large language model in December 2024, rivaling US models like OpenAI's GPT and Anthropic's Claude at a fraction of the cost.
- "Fable," "Mythos," and "ChatGPT 5.6" are speculative/satirical names the author uses—these are not real products. The author is using fictional names to mock the narrative that advanced AI needs strict government control.
- The Washington Post report (linked) discusses actual US government proposals to license and restrict access to cutting-edge AI models, citing national security concerns.
- The core tension: US AI companies (OpenAI, Anthropic) advocate for safety regulation, which critics say would lock out cheaper, open-source competitors—especially Chinese ones—under the banner of security.