GuShiio.com ① The most frequently asked financial questions DeepSeek include investment and financial management, stock/fund purchases, mortgages, inflation, financial scams, retirement/education funds, credit management, etc.;
②DeepSeek proposes common cognitive blind spots among users, such as underestimating the power of compound interest, overestimating short-term benefits, and ignoring hidden costs, and recommends knowledge reserve, tool practice and professional consultation.
DeepSeek’s ecosystem continues to expand. Recently, with WeChat announcing its access to the DeepSeek-R1 model, the prediction that “1.3 billion users will move into DeepSeek” seems to come true. According to data released by the AI product list, in January 2025, the number of monthly active users of DeepSeek’s domestic APP reached 33.7 million; the monthly visits to DeepSeek’s website exceeded 256 million, which is more than seven times the monthly visits to Doubao and Kimi.
With such a large number of users, what are the most frequently asked financial questions DeepSeek so far? The Financial Times reporter asked DeepSeek this question after choosing to use “Deep Thinking (R1)”.
It gives seven types of financial questions that are frequently asked. The evaluation of these questions reflects the public’s general concerns about financial management, investment and risk management, mainly focusing on investment and financial management, purchase of stocks/funds, mortgage loans, inflation, financial scams, Retirement/education funds, credit management, etc., and gives his own thoughts, and also lists “common cognitive blind spots among users” and “recommended actions” as derivative answers. Specifically, it includes:
1. How to start financial management/investment?
2. Should I buy stocks/funds/cryptocurrencies now?
3. Mortgage/car loan decision issues
4. How to deal with inflation?
5. Identification of financial scams
6. Retirement/education planning
7. Credit management issues
DeepSeek proposes three common cognitive blind spots for users. One is to underestimate the power of compound interest, such as a fixed monthly investment of 1000 yuan, annualized by 8%, which can reach 1.5 million yuan in 30 years; the other is to overestimate short-term returns, such as the long-term annualization of A shares. About 7%-10%, but annual fluctuations often reach ±20%; the third is to ignore hidden costs, such as the erosion effect of fund subscription fees and management fees on long-term returns.
DeepSeek gave three suggestions: one is to reserve knowledge and recommend Coursera’s “Personal Financial Planning” or “Financial Markets” courses; the other is to practice tools and use simulated stock trading apps (such as Investopedia Simulator) to test the water; and the third is to provide professional consultation, and complex tax/inheritance issues need to be contacted by CFP licensees.
A financial person told the Financial Times reporter that the above answer is certain professional. At the end of the answer, DeepSeek also specifically stated that financial decisions need to be highly personalized, and the above suggestions are for reference only.
It is worth mentioning that DeepSeek-R1 ‘s strong logical thinking ability when answering the question impressed reporters. Regarding the questions raised by reporters, it first analyzed and needed to recall the previous interactions with other users; next, it analyzed why these questions were frequently asked; then, it expressed the need to confirm whether some questions became popular within a specific period of time; In addition, it concludes that “users may lack basic financial knowledge, causing them to repeatedly ask about some basic concepts”, and judges that it needs to be based on the user’s knowledge level in order to provide appropriate explanations and avoid using overly professional terms, or conversely, overly simplified answers.
In addition, DeepSeek-R1 also explains that users in different regions may have different common questions that need to be considered. For example, users in the United States may be more concerned about 401(k) or IRA retirement plans, while users in China may be more concerned about the choice of housing provident fund or wealth management products. Therefore, regional differences should be taken into account when answering, or more general suggestions should be provided when the user’s region cannot be determined.
DeepSeek-R1 judges that financial issues often involve sensitive information, such as personal income, debt status, etc. Users may not provide detailed information when asking questions, so they are required to emphasize general suggestions when answering and advise them to consult a professional financial consultant to get personalized guidance. We must also pay attention to compliance issues, ensure that answers comply with relevant laws and regulations, and do not provide specific investment advice. Especially when it comes to individual stock recommendations or market forecasts, we need to avoid legal risks and remain neutral and objective.
The reporter learned that in the financial field, DeepSeek has been piloted to varying degrees in scenarios such as intelligent customer service assistants, financial marketing, product consulting, securities investment research, risk control, and credit review. However, no complex transaction areas have been involved for the time being.
“AI technology is undergoing a transformation from ‘instrumental application’ to ‘systematic reconstruction’. Technological breakthroughs such as DeepSeek will further promote the in-depth application of AI in finance, medical, manufacturing and other industries, and accelerate multimodal AI integration (text, image, voice collaboration) and the development of autonomous decision-making agents. In the financial industry, AI risk control systems can identify abnormal transactions in real time, reduce fraud risks, and reduce manual review costs.” Wang Yi, chief economist of Great Wall Securities, told the Financial Times reporter that however, the popularization of AI still faces standardization and ethical issues, especially in medical, financial and other fields, where standards for data privacy protection and algorithm interpretability need to be established to prevent technology abuse.
