The Impact of Artificial Intelligence on the Financial Industry: From Initial Anxiety to Rational Application

The emergence of ChatGPT has triggered a wave of anxiety in the financial industry. As a sector that is deeply rooted in technology, the financial world is worried about being left behind in the rapidly evolving era of artificial intelligence. This anxiety has spread to every corner, with discussions about large model technology even being heard among finance professionals in temples.

However, over time, this anxiety gradually subsided, and people's thinking became clearer and more rational. The financial industry's attitude towards large models has gone through several stages: initial concerns and anxieties, followed by proactive actions, then rational thinking after encountering practical difficulties, and now it has entered a stage of selectively trying out validated scenarios.

It is worth noting that many financial institutions have begun to place strategic importance on large model technology. According to statistics, at least 11 banks among A-share listed companies have explicitly stated in their latest semi-annual reports that they are exploring the application of large models. From recent actions, these institutions are engaging in deeper thinking and planning from a strategic and top-level design perspective.

From Enthusiasm to Rational Return

At the beginning of the year, when ChatGPT was first launched, the financial sector had a limited understanding of large models, despite their enthusiasm. Some large banks took the lead and began various marketing promotions. At the same time, the technology departments of some leading financial institutions actively engaged with large technology companies to discuss the construction of large models.

However, after May, due to factors such as the shortage of computing power resources and high costs, many financial institutions began to shift their focus from simply building models to emphasizing application value. Enterprises of different sizes also started to adopt different strategies: large financial institutions tend to introduce industry-leading foundational large models and build their own enterprise large models, while small and medium-sized financial institutions consider introducing various large model services on demand.

Nevertheless, due to the high demands for data compliance, security, and reliability in the financial industry, the implementation progress of large models in this field has actually been slower than expected at the beginning of the year. To address various obstacles in the implementation process, financial institutions are adopting multiple approaches, including building their own computing power and hybrid deployment. At the same time, an increasing number of financial institutions are also beginning to strengthen their data governance efforts.

Entering from the peripheral scene

In the past six months, financial institutions and large model service providers have been actively exploring various application scenarios, including smart office, intelligent development, smart marketing, intelligent customer service, smart investment research, and intelligent risk control. Each financial institution has rich ideas regarding large models.

However, in practice, the general consensus is to start with internal applications and then gradually expand to external ones. This is because the current large model technology is still not mature, and the financial industry is a field with strong regulation and high security requirements. Therefore, many institutions choose to start with relatively easy-to-implement scenarios such as code assistants and customer service assistants.

It is worth noting that these widely implemented scenarios are not yet the core applications of financial institutions. For large models to truly penetrate the business layer of the financial industry, certain time and technological breakthroughs are still required.

At the same time, some reforms at the top-level design are underway. More and more financial institutions are beginning to build multi-level system frameworks based on large models, using large models as the central hub while integrating traditional models and adopting multi-model strategies to optimize results.

The talent gap remains large

The application of large models has begun to challenge and transform the personnel structure of the financial industry. Some traditional positions face the risk of being replaced, but at the same time, new opportunities and demands have been created.

Many financial institutions hope to improve the quality of service and work efficiency of their employees through large models, rather than simply replacing human labor. However, a major challenge currently faced by the industry is the severe shortage of talent related to large models.

Financial institutions need composite talents who understand both finance and AI, especially in the development of industry-specific or enterprise-level models. To this end, some institutions have begun to take action, such as designing training courses and establishing joint project teams, to enhance the relevant capabilities of internal staff.

With the continuous development and application of large model technology, the personnel structure of financial institutions will also undergo adjustments and changes. Those who are proficient in using large model technology are more likely to stand out in this rapidly changing environment.