IDG、哈勃、元禾，联手投了复旦知名教授邱锡鹏丨投中嘉川

Core Viewpoint - The article discusses the recent funding round for MOSS Intelligence, a company founded by renowned AI professor Qiu Xipeng, highlighting its advancements in speech AI technology and the significance of its latest product, MOSS-Speech, which marks a new phase in voice interaction technology [5][15]. Company Overview - MOSS Intelligence, established in November 2024 in Shanghai, aims to promote the industrialization of multimodal interaction and embodied intelligence, focusing on creating AI systems that better understand human interactions [7]. - The company has attracted top-tier investments from IDG Capital, Huawei Hubble, Yuanhe Holdings, and Xinglian Capital during its angel round [5]. Key Technology and Products - MOSS Intelligence specializes in the development of large speech models, with its latest release, MOSS-Speech, representing a significant advancement in AI voice interaction, allowing for direct voice-to-voice communication without text intermediaries [15][16]. - MOSS-Speech captures non-verbal signals such as tone, emotion, and laughter, making machine speech more natural and human-like [16]. Founder's Background - Qiu Xipeng, a professor at Fudan University and a prominent figure in the AI field, co-founded MOSS Intelligence. He is known for his influential work in natural language processing and deep learning, including the widely regarded textbook "Neural Networks and Deep Learning" [8][9]. - Qiu's team developed the open-source Chinese language model MOSS, which filled a significant gap in the Chinese AI landscape following the release of ChatGPT [9][10]. Leadership Team - The CEO of MOSS Intelligence, Li Shiming, is a graduate student of Qiu Xipeng and has made notable contributions to the field, including the development of SpeechGPT, a pioneering voice interaction model [14]. - Li has also led the creation of a cross-disciplinary collaboration platform, demonstrating a commitment to transforming cutting-edge technology into socially valuable products [14]. Future Aspirations - The company aims to develop a multimodal large model and content creation platform that rivals Google's Gemini Pro, with a focus on high-performance audio and video understanding and generation capabilities [15]. - Qiu Xipeng envisions MOSS as a foundational step towards achieving general artificial intelligence, suggesting that the realization of such technology may be within the next 5 to 10 years [16].