10 Questioning Tips for DeepSeek 49

To ask good questions, you only need to follow two rules—input and output. Since DeepSeek has exceptionally strong reasoning capabilities, all you need to do is provide sufficient background information for your question and specify the rules that DeepSeek must follow when giving its answer. These rules relate to language style, and you can use these prompt to make demands of DeepSeek. For example, phrases like "speak plainly," "use language a middle school student could understand," or "make it understandable to non-professionals" can dramatically unlock DeepSeek's capabilities. Below are 10 question-asking techniques recommended for DeepSeek:

1. State Your Needs Clearly and Specifically

Example: ❌ "Explain artificial intelligence"

✅ "Please explain the working principles of generative ai in plain language, and provide examples of its applications in the healthcare field"

2. Break Questions into Steps

Example: ❌ "How do I develop an APP?"

✅ "Please describe the basic process of mobile app development in 5 steps, with emphasis on the key tasks of the requirements analysis phase"

3. Limit the Scope of the Answer

Example: ❌ "Introduce Python"

✅ "Please summarize the three core application scenarios of Python in data science within 200 words, and recommend the corresponding commonly used libraries"

4. Specify the Answer Format

Example: ❌ "Recommend a book list"

✅ "Please recommend 5 introductory philosophy books that improve logical thinking, formatted as [Title | Author | Reason for Recommendation]"

5. Request Confirmation and Verification

Example: ❌ "Help me write a resume"

✅ "Please first list the job-seeking information items you need to know, and continue generating the resume content after I provide the details"

6. Compare from Multiple Perspectives

Example: ❌ "What is blockchain"

✅ "Please compare the similarities and differences between blockchain and traditional databases from three dimensions: technical principles, application scenarios, and legal regulation"

7. Provide Context

Example: ❌ "This code has a problem"

✅ "Below is a Python data cleaning code (code attached), which throws a ValueError when run. Please analyze the possible causes and provide solutions"

8. Set Constraints

Example: ❌ "Plan a marketing scheme"

✅ "Please design a summer promotion plan for a small milk tea shop with a budget of 50,000 yuan, covering 3 online channels and including ROI estimates"

9. Use Professional Terminology

Example: ❌ "How to analyze stocks"

✅ "Please use the DDM model to analyze the fair value of Kweichow Moutai (600519), and include explanations of the necessary parameter assumptions"

10. Progressive Follow-Up Questions

Examples:

Q1: "Briefly describe the three main learning paradigms of machine learning"

Q2: "Please focus on explaining the applicable scenarios of the Random Forest algorithm in supervised learning"

Q3: "Regarding your answer to Q2, please supplement it with a specific application case of this algorithm in financial risk control"

DeepSeek Usage Tips:

• Structured Expression: Use a "Background + Requirement + Constraints" question structure

• Specific Requirements: Quantify indicators (word count/quantity) and clarify comparison dimensions

• Interactive Optimization: Follow up with "Please add..." / "Could you elaborate..."

• Professional Level Adaptation: Adjust the use of terminology according to your own knowledge level

By applying the 10 question-asking techniques above, you can essentially obtain the information you need more accurately. It is recommended that you flexibly combine and use them based on your actual scenarios.