On December 17, Gelonghui reported that Xinhuanet Co., Ltd. (603888.SH) stated on the interactive platform that on August 23, during the 2024 Solar Island Entrepreneurs Conference, it launched the "Xinhuanet AIGC Application Enablement Platform." This platform is an industrial infrastructure platform promoted by Xinhuanet in the field of AI, providing one-stop exclusive large model training and application development services, enabling clients to obtain reliable, flexible, secure, and low-cost integrated intelligent computing services through a "cloud-edge-end collaboration" architecture. The platform boasts four core capabilities: elastic and reliable green computing power, professional and authoritative industry corpora, secure and controllable model applications, and scenario-oriented training platform, assisting enterprises in agilely building exclusive large models. Meanwhile, Xinhuanet jointly published the "Multimodal Publishing Corpus Production and Intelligent Service System" and the "High-Performance Elastic Computing Power Package," two major industry products developed based on the platform, which will enhance innovation supply in terms of "corpora" and "computing power," further promoting the popularization and application of intelligent computing services. On the morning of September 20, during the Beijing Cultural Forum, Xinhuanet prominently released the "AIGC-Safe Comprehensive Service Platform for Universal Content Security and Model Security." This platform encompasses four core technical capabilities: at the model level, introducing the first domestic large model for forgery detection, and setting up an AI offense-defense platform and data generation foundation to ensure continuous model updates; at the data level, leveraging the advantages of hosting China's Internet Joint Rumor Refutation Platform, accumulating a massive database of rumors and a repository of characteristics of harmful information, focusing on reserving and training against forgery algorithms and synthetic data; at the algorithm level, ensuring efficient and precise deployment of applications across various environments and scenarios through deep algorithm optimization; at the hardware level, adapting to domestic chips, optimizing software-hardware collaboration, and network acceleration, developing integrated machine products that ensure the safety of trusted innovation. AIGC-Safe has broad application prospects in various scenarios, which can be widely used in detecting false information, forgery detection of specific individuals, AI face-swapping, voice synthesis fraud detection, account face-swapping detection, news authenticity detection, academic cheating detection, ticket verification, public display content safety control, public opinion risk control, and copyright risk control, among many areas, solving the practical problems accompanying the development of AI in real scenarios. On December 3, Xinhuanet and Beijing Kingsoft Office Software, Inc. signed a cooperation agreement during the 2024 Boao Forum for Entrepreneurs series of events; in the future, functionalities such as "AIGC Detection," "Original Certification," and "Copyright Operation" will be embedded in the "WPS Office" software, jointly creating a "Smart Barrier" to safeguard the booming development of the academic industry.
新华网(603888.SH):将强化“语料”和“算力”上的创新供给,进一步推动智算服务的普及与应用
Xinhuanet Co., Ltd. (603888.SH): Will strengthen innovative supply in "corpora" and "computational power", further promoting the popularization and application of intelligent computing services.
The translation is provided by third-party software.
The above content is for informational or educational purposes only and does not constitute any investment advice related to Futu. Although we strive to ensure the truthfulness, accuracy, and originality of all such content, we cannot guarantee it.