Generated Knowledge Prompting
Knowledge-based responses through context generation
Generated Knowledge Prompting improves AI responses by first generating relevant background knowledge and then using it to answer the actual question.
How does it work?
This technique consists of two steps:
- Knowledge Generation: The AI creates relevant information about the topic
- Finding the Answer: Based on this knowledge, it answers your question
Practical Examples
Scientific Explanations
Step 1 - Generate knowledge about photosynthesis:
"Explain the key aspects of photosynthesis"
AI Response:
"Photosynthesis is the process where plants convert light energy into chemical
energy. Chlorophyll in the leaves absorbs light, which together with water
and CO2 is converted into glucose and oxygen."
Step 2 - Answer specific question:
"Based on this knowledge: Why are most leaves green?"
Answer:
"Leaves are green because they contain chlorophyll. Chlorophyll efficiently
absorbs red and blue light but reflects green light, which is why we
perceive leaves as green."
Business Decisions
Step 1 - Generate market knowledge:
"What are the current trends in e-commerce for sustainable products?"
AI generates information about:
- Growth of the sustainable products market
- Consumer behavior and preferences
- Successful business models
Step 2 - Use this knowledge:
"As a traditional retailer, should we invest in sustainable products?"
Well-founded answer based on the generated market data
Technical Problems
Step 1 - Generate technical knowledge:
"Explain the most common causes of slow website performance"
AI lists:
- Large, uncompressed images
- Too many HTTP requests
- Missing browser caching
- Slow server response times
Step 2 - Solve specific problem:
"Our e-commerce site loads very slowly. What should we check first?"
Targeted recommendations based on the generated knowledge
Variants of the Technique
Automatic Knowledge Generation
"Answer this question and first generate relevant background knowledge:
What factors influence employee satisfaction in remote teams?"
Structured Knowledge Generation
"Generate the following knowledge about project management:
1. Most important methods
2. Typical challenges
3. Best practices
Then use this knowledge to explain how to optimally structure an
IT project with 50 employees."
Multiple Knowledge Generation
"Generate three different perspectives on this topic:
1. Technical view
2. Business view
3. User view
Topic: Implementation of a new CRM software"
Best Practices
Target
Ensure Relevance
Make sure the generated knowledge fits the question
Layout
Define Structure
Specify what kind of knowledge should be generated
Shield
Verification
Have the AI verify the generated knowledge for accuracy
Advanced Applications
Domain-specific Knowledge
"As a financial expert, generate knowledge about:
- Current interest rate developments
- Inflation impacts
- Investment strategies in volatile markets
Use this knowledge: How should a conservative investor
structure their portfolio in 2024?"
Creative Applications
"Generate knowledge about successful sustainability advertising campaigns.
Based on this, develop three campaign ideas for our
new environmentally friendly product."
Problem Solving
"Gather information about typical causes of high employee turnover.
Then analyze our situation:
- 30% turnover in the last year
- Mainly young employees affected
- IT industry
What are the most likely causes?"
When to Use This Technique
✅ Ideal for:
- Complex questions requiring background knowledge
- Situations where context is important
- Decision-making with multiple factors
- Explanations of scientific or technical concepts
❌ Less suitable for:
- Simple factual questions
- When you already have all necessary information
- Time-critical requests (two steps take longer)
Practical Templates
Business Analysis
1. "Generate knowledge about [industry/market/technology]"
2. "Based on this, analyze [specific situation]"
Technical Solution
1. "Explain the basics of [technology/concept]"
2. "How can this help with [specific problem]?"
Strategy Development
1. "What are best practices for [area]?"
2. "Develop a strategy for [specific challenge]"
Pro Tip: Combine Generated Knowledge with Chain-of-Thought for even deeper analyses!
Exercise
Try Generated Knowledge Prompting for:
- A technical challenge in your company
- A strategic business decision
- Explaining a complex concept to laypeople
Next Step: Discover Self-Consistency for more robust answers through multiple perspectives.