وكالة صدى نيوز
وكالة صدى نيوز
Ready to play
Ready to play
A review of over 80 Chinese research papers and patents has revealed that military researchers in China are using outputs from advanced American AI models, such as GPT-3.5, to train local AI systems aimed at enhancing China's defense capabilities. Chinese military institutions rely on a technique called "model distillation," which leverages the outputs of large AI models to train smaller models that can be operated locally. This approach accelerates the development of specific capabilities despite U.S. restrictions on exporting advanced technologies. American accusations allege that Beijing is using this method to extract U.S. AI capabilities that are threatened by export controls; however, China denies these claims and accuses Washington of trying to dominate the field. Meanwhile, research suggests that China is seeking to develop its capabilities independently through these techniques.
Notice: This Is an AI-Generated Summary
Comments (0)