top of page
Writer's pictureGJC Team

Prompt engineering - top tips

Updated: Oct 29


prompt engineering

What is prompt engineering and why is it so important?


Prompt engineering is the practice of crafting inputs—called prompts—that guide AI language models to produce desired outcomes. By carefully designing these prompts, users can ensure more accurate, relevant, and insightful results. As AI systems grow more powerful, understanding how to effectively communicate with them becomes crucial. Mastering prompt engineering enhances the interaction between humans and AI, leading to better, faster, and more reliable results.


What is prompt engineering?


Prompt engineering involves creating specific inputs that help AI understand the user’s request and generate relevant responses. By strategically designing prompts, users can ensure the AI operates within a clearly defined scope, improving communication and output quality. The practice blends creativity with an understanding of AI’s capabilities and limitations.


Why is prompt engineering important?


Prompt engineering directly impacts the effectiveness of AI outputs. Well-structured prompts lead to more helpful responses, making interactions with AI smoother and more productive. This skill is essential for professionals and everyday users alike, allowing faster problem-solving, more precise insights, and improved decision-making.


Other benefits of using prompt engineering include: faster response time that allows users to extract information more efficiently; higher accuracy - better designed prompts leading to more accurate and relevant responses; managing complexity - breaking down queries into simpler components improves AI’s ability to address it; better user experience because there are more concise and coherent and higher-quality outputs; and supporting innovation by refining how users interact with AI.


Good practice in prompt engineering


Effective prompt engineering requires understanding how to communicate clearly with AI systems. Following these best practices could improve the quality and relevance of AI responses:


  1. Choose words carefully: AI tools rely on precise language, so avoid slang words, ambiguity, or metaphors unless you’re aiming for creative, unpredictable results. Clear, direct wording will produce the most consistent and reliable outputs.

  2. Break down complex tasks: Divide questions into smaller, manageable parts. This helps the AI focus on each step, leading to more thorough and accurate responses. For example, instead of asking for a comprehensive report, break the request into sections, like "Outline the causes of World War 2" followed by "discuss the economic conditions in pre-war Germany."

  3. Understand the model’s limitations: Recognise that AI has limitations. It may not understand context perfectly and may lack real-time data processing. Set realistic expectations and adjust your prompts to avoid incomplete or inaccurate results.

  4. Provide the desired output format: Specify how you want the information to be presented. Do you need a summary, a list, or a detailed report? Defining the format ensures the AI tailors its response to your needs.

  5. Being specific: Provide detailed and clear prompts that include all necessary context. Specify format, length, tone, and other details to guide the AI toward the most accurate result. For example, instead of asking, "Tell me about the history of Europe," specify, "Explain the impact of World War 2 on the development of the European Union."

  6. Encourage detailed reasoning when required: Using a "chain of thought" approach can prompt the AI to explain its reasoning process. This is especially useful when the reasoning behind an answer is as important as the answer itself.

  7. Assign a 'persona' or frame of reference: Providing the AI with a specific role or context, such as “Respond as an economic expert,” improves the accuracy and relevance of the response. This technique helps the AI focus on the appropriate tone and expertise level.

  8. Focus on what to do, not what not to do: Frame instructions positively. Telling the AI what to achieve is more effective than explaining what it should avoid. Positive framing helps reduce confusion and ensures the AI is focused on the correct tasks.

  9. Experiment with prompting: AI systems can interpret prompts differently depending on their structure. Experimenting with variations in phrasing and format can yield better results. Regularly testing different approaches helps users learn how AI interprets inputs and refine their prompt design over time.

  10. Understand the outcome sought: Clearly define the goals before interacting with the AI. Knowing what you want to achieve helps structure prompts more effectively, guiding the AI to produce relevant results that align with your expectations.

  11. Don't make the prompt too long: Long prompts can confuse AI systems, making them struggle to prioritise and organise the essential elements of a request. Keep prompts concise and to the point, only as long as necessary to convey all required details.

  12. Use open-ended questions: AI systems thrive on open-ended prompts, which allow for more detailed and creative responses. Instead of yes-or-no questions, frame your prompts to encourage the AI to explore a topic in depth. For instance, rather than asking, “Is the sea blue?” ask, “Why does the sea appear blue when it is a sunny day?”

  13. Include context: Providing context helps the AI tailor its response to your intended audience. A prompt such as, "Explain the economy of Australia" for a high school student versus for another international business audience.

  14. Set length parameters: AI models can struggle with precise word or character limits, you can request general output guidelines, such as “Provide a short summary” or “Write a detailed, 400-word report.” These suggestions help the AI produce responses that align with your expectations.

  15. Avoid conflicting requests: Be consistent in your prompts and avoid contradictory instructions as these can confuse the AI and lead to unclear results.

  16. Use punctuation to clarify: Just as punctuation helps humans parse text, it can also guide AI in understanding complex prompts. Proper use of commas, quotation marks, and line breaks can help the AI process and deliver clearer, more organised responses.


Conclusion


Prompt engineering is an essential skill for leveraging AI's full potential. By following these best practices—such as being specific, breaking down complex tasks, and understanding the AI’s limitations—users can ensure more effective and accurate results. Whether for content creation, data analysis, or problem-solving, strong prompt engineering transforms the way we interact with AI, leading to faster, better, and more innovative outcomes.




GJC


1 Comment


Guest
Oct 23

What is Google Gemini used for?

Like
bottom of page