An industry insider revealed that Lei Jun, recognizing Xiaomi's late entry into the realm of large AI models, personally embarked on a recruitment drive to lead Xiaomi's large model endeavors, offering a salary in the vicinity of ten million yuan.

Luo Fuli, a post-95s AI engineer, stands out as a key developer behind the open-source large model DeepSeek-V2. Her prowess and experience in the AI field make her a highly coveted talent for Xiaomi.

About Luo Fuli

Born in 1995 to an ordinary family in Sichuan, Luo Fuli graduated with a major in computer science from Beijing Normal University and went on to pursue graduate studies at the Institute of Computational Linguistics at Peking University.

During her tenure at Peking University, Luo Fuli authored eight papers at the prestigious AI conference ACL, two of which she led, creating a buzz in the international academic sphere. Her research on word sense disambiguation methods and text style transfer frameworks significantly advanced natural language processing technology.

After obtaining her master's degree, Luo Fuli initially joined Alibaba's DAMO Academy, spearheading the development of the multilingual pre-trained model VECO. She subsequently moved to Quantopian to engage in deep learning-related strategy modeling and algorithm research.

On Bilibili, Luo Fuli goes by the username "I'm not an AI Lolita," with a profile stating, "A post-95s AI researcher, graduated from Peking University with a master's degree, currently working at Alibaba's DAMO Academy." (Note: Subsequent job updates have not been reflected on her profile.)

In 2022, she joined DeepSeek as a deep learning researcher and contributed to the development of the MoE large model DeepSeek-V2.

Recently, it was disclosed that Xiaomi's founder, Lei Jun, offered Luo Fuli a multi-million salary to join Xiaomi AI Lab. Her work history on Baidu Baike has since been updated accordingly.

May Lead Xiaomi's Large Model Team

Xiaomi has been deeply entrenched in the field of artificial intelligence for several years, establishing various professional teams, including Xiaoai Assistant, autonomous driving, and the AI Lab. In April of last year, Xiaomi officially formed its large model team within the AI Lab, headed by Luan Jian. According to Jiemian News, the team was initially equipped with 6,500 GPU resources and is actively constructing a GPU cluster with 10,000 cards to fully propel the development of large AI models.

Xiaomi's consistent investment is beginning to bear fruit, with the Xiaomi 15 exhibiting impressive performance in basic AI functions. However, there is still room for improvement compared to competitors in more advanced applications, such as multimodal interaction and cross-application collaboration.

Meanwhile, brands like Honor, Huawei, OPPO, and vivo have elevated AI technology to a strategic level. Li Xiang, CEO of NIO, even announced that the company is transforming from a car manufacturer into an AI company. ByteDance's AI capital expenditure for 2024 stands at a staggering 80 billion yuan.

In terms of potential, Xiaomi boasts one of the world's richest AI entry points and application scenarios, with over 861 million IoT devices connected to its AIoT platform. With the support of AI models, Xiaomi's mobile phones, automobiles, and smart home products can achieve more efficient intra-scene and cross-scene multi-device collaboration, elevating the smart experience to a new level and providing robust support for its "people-car-home ecosystem" strategy.

Upon joining Xiaomi, Luo Fuli is anticipated to take on a pivotal role in Xiaomi AI Lab, leading the large model team. Xiaomi hopes that her arrival will catalyze the implementation of its AI strategy and broaden its business horizons and market space.