Emphasizing the Importance of Open Source, Zuckerberg Calls on the U.S. Government to Reassess AI Model Distillation Policies

Mark Zuckerberg, CEO of Meta, emphasized the importance of open-source artificial intelligence in a post on the company's official website on the 10th and called on the U.S. government to "re-examine policies regarding model distillation and the use of training data," in order to protect the principle that allows models to learn from other models during the development process. In an article titled "The Future Belongs to Everyone," Zuckerberg stated that the ability of models to learn from other models is a crucial principle for the functioning of the open-source ecosystem. All AI models are derived from human knowledge. While some attempt to describe distillation as harmful, he maintains that the principle of "learning from anything observable" should be protected. Model distillation is a foundational technology in the development of AI models, typically referring to the efficient transfer of knowledge from larger AI models to lightweight models. As competition in the development of generative AI intensifies, model distillation has also become a topic of industry focus and controversy. Zuckerberg noted that U.S. labs face more restrictions on training data. If the U.S. wishes to maintain a long-term lead in open-source models, it should reduce these additional barriers. Limiting foreign open-source models is not an effective solution. He believes that restricting individuals and businesses from using leading open-source models, regardless of their origin, would diminish the quality of AI available to users and further centralize AI rather than empowering people. Zuckerberg also stated that Meta will continue to strongly support open-source initiatives, including open-source AI models. As the company's superintelligence lab comes online, it will resume the release of open-source models.