AI prompts
Prompt engineering
Getting the best out of AI language models is like giving them the right instructions. Prompt engineering is the art of crafting those instructions - the questions or commands - to get the most accurate, clear, and relevant answers possible. It's about fine-tuning what you tell the AI to make sure it delivers what you need.
The CLEAR framework
The CLEAR framework, is a standardized method for creating effective prompts for AI language models. The following elements ensure users craft clear and well-structured prompts that guide the AI towards generating high-quality, relevant, and contextually appropriate content. The CLEAR Framework emphasizes that prompts should be:
- Concise: brief and to the point, focusing only on necessary information for the desired AI response.
- Logical: clear and coherent, ensuring a smooth flow of information and instructions for the AI model.
- Explicit: provide specific details regarding the desired output format, content, and scope, leaving no room for ambiguity in the AI's generation process.
- Adaptive: there is need for flexibility in prompts. Users can adjust them based on the AI's response and the evolving needs of the task.
- Reflective: users should continually assess and improve their prompts based on the AI's output. This ensures ongoing optimization for the best possible results.
Consult the following table that provides examples which illustrate how to apply the CLEAR framework in practice.
| CLEAR Framework element |
Prompt |
Improved Prompt |
|
Concise
brevity and clarity in prompts
|
Can you provide me with a detailed explanation of the process of photosynthesis and its significance? |
Explain the process of photosynthesis and its significance |
|
Logical
structured and coherent prompts
|
Give me the steps of the scientific method |
Describe the steps in the scientific method, starting with forming a hypothesis and ending with drawing conclusions |
|
Explicit
clear output specifications
|
What are some renewable energy sources? |
Identify five renewable energy sources and explain how each works |
|
Logical
structured and coherent prompts
|
Describe the history of computers |
Explain the development of personal computers from the 1970s to the 1990s |
|
Reflective
continuous evaluation and improvement of prompts
|
After acquiring an AI-generated list of strategies for effective time management, evaluate the relevance and applicability of each strategy. Consider the target audience's needs, and use this information to tailor future prompts to generate content that better addresses specific challenges or contexts. |
The technical aspects of prompts
We must also be aware of the technical aspects of prompts that can influence AI-generated content in addition to the CLEAR principles. Among the most important technical factors are:
-
Building Blocks of Text: Think of AI language models like chefs – they work with ingredients to create something new. These ingredients, called tokens, are the basic units of text, like words or even single characters. Keeping prompts concise ensures the "chef" has all the ingredients to create a complete and relevant response.
-
Turning Up the Creativity Dial: Imagine a dial that controls how surprising the AI's response will be. A lower setting (like "low heat") creates focused and predictable results, while a higher setting (like "high heat") produces more creative and varied responses. You can experiment with this setting to find the perfect balance between fresh ideas and clear communication.
-
Focusing on the Most Likely Options: Another setting, called "top-p," allows librarians to fine-tune the randomness of the AI's response. This works by prioritizing the most likely words based on a set probability. By adjusting this setting, you can steer the AI towards more focused or diverse answers, depending on the task at hand.