Recently, the first FCon Global Financial Technology Conference was grandly opened in Shanghai, and Fei Haojun, Chief Algorithm Scientist of Qifu Technology, was invited to become the producer of the "Financial Industry Big Model Application" sub forum. He planned and organized multiple experts from industry institutions such as Everbright Trust, Jianxin Jinke, Ant Group, and Zhongguancun Kejin to share their ideas.
Fei Haojun said in the opening introduction: "The explosion of ChatGPT this year has accelerated the process of empowering the financial industry with big models. The data-driven process of the financial industry has given it an inherent advantage in empowering big models, and the complex and professional knowledge system of the financial industry also provides a unique historical mission for the application of big models. The industrial transformation of big models is destined to be a long-lasting battle, which cannot be separated from every down-to-earth application practice."
Dr. Zhu Shihu, Deputy General Manager of the Information Technology Department and General Manager of the Data Center of Everbright Trust, pointed out that the big model is an improvement in productivity, and the implementation path in the financial industry relies on big cooperation and innovation. By integrating big data and collaborating with large computing power, we can fine tune models in vertical fields and create lightweight inference models with small-scale computing power. In terms of big innovation, the current application of big models in the financial field is mainly focused on intelligent customer service, intelligent operations, intelligent office, and other fields. In the future, they should be more applied in risk management, capital management, and regulatory technology.
Liu Dongdong, General Manager of the Artificial Intelligence Engineering Department at Jianxin Jinke Basic Technology Center, stated that the emergence of ChatGPT4.0 has redefined the relationship between humans and machines, and is considered the fourth industrial revolution not seen in 250 years, which will inevitably lead to industry reshuffle. This tool upgrade is not a short-term technological hotspot. The big model is the same fundamental technology as the chip, which not only affects the company's competitiveness, but also will inevitably affect the country's competitive discourse power.
Xu Wanqing, a senior technical expert at Ant Group, emphasized the practice and application of Ant Financial's big model in investment research and financial services scenarios. Facing the increasing competitive pressure in the wealth service industry for high net worth clients, it is particularly important to enhance the professional ability and business efficiency of financial advisors. This can not only improve service quality, but also expand service scope.
Dr. Zhang Jie, Vice President of Zhongguancun Science and Technology, believes that big models have become an important driving force for innovative development in the financial industry. This is not only a technological innovation in the financial industry, but also an important engine for promoting the upgrading and transformation of financial business. The emergence of applications based on large models is bound to bring better processing methods for massive data and complex business scenarios in the financial industry. However, the financial industry has its unique industry attributes and regulatory requirements, and many challenges in the application of large models in the financial field are also inevitable.
Fei Haojun proposed that fintech companies are the first battlefield for the implementation of financial big models. In the practice of implementing financial models, Qifu Technology has achieved remarkable business results through the powerful empowerment of AIGC. In the telemarketing system, semantic analysis and lead mining help improve the accuracy of telemarketing lead recognition by up to 98%, while increasing conversion rates by over 5%. In the intelligent marketing process, approximately 70% of image materials are generated by AIGC, and it is planned to annotate and rate the materials in multiple dimensions through a large model to achieve optimization of advertising placement. In the call quality inspection process, the automated quality inspection of the large model achieved 100% coverage, increasing the detection rate by 15%. In the scenario of speech robot script generation, the quality rate of generated scripts has reached 70%. Fei Haojun said, "The development of the financial big model requires the joint cooperation of financial institutions and fintech companies." (Economic Daily reporter Peng Jiang)