Nathan Lambert Questions How Proposed U.S. Disclosure Rules Would Define “Chinese Models”
Nathan Lambert questioned how to define a “Chinese model” in response to a proposal reportedly under discussion by the U.S. government that would require companies to publicly disclose their use of Chinese open-weight AI models. He cited the Thinky model distilled from Kimi, Composer after continued pretraining, Nemotron trained using large amounts of synthetic data from DeepSeek, and a Nous model fine-tuned around “American” values, illustrating why model provenance cannot be neatly categorized.
Why read this
AI score 56
Models may combine knowledge distillation, continued pretraining, synthetic data, and values-based fine-tuning, creating practical definitional challenges when classifying them by country.
Topics
View original postSITUATION DETECTED: The White House and Treasury are discussing a proposal that would require US companies to publicly disclose the use of Chinese open-weight AI models on their websites, aiming to discourage their use through public pressure, per Reuters.
View original
SITUATION DETECTED: The White House and Treasury are discussing a proposal that would require US companies to publicly disclose the use of Chinese open-weight AI models on their websites, aiming to discourage their use through public pressure, per Reuters.
Body
This is stupid. What counts as a chinese model? What about Thinky's model distilled from Kimi? What about composer, continued pretraining? What about Nemotron, using tons of DeepSeek synthetic data? What about that silly nous finetune for "american" values?
AI Trends