Director, Learning Design & Development| PMIAsheville, NC, United States
With Generative AI, iteratively refining and optimizing prompts can lead to better AI-generated results. This may involve adjusting the specificity or clarity of the prompt to increase relevance and accuracy of results.
What examples do you have of how improving a prompt drastically changed the output quality? What specific changes did you make that led to the improvement?
When creating a prompt, it's crucial to clearly define the role, the context in which the role operates (such as location), the tasks the role needs to accomplish, and the format of the response. This is known as the R-T-F (Role-Task-Format) framework. Eliminating ambiguity is essential to avoid hallucinations in the AI's responses. Saving Changes...
ROBSON BANDEIRAADVISOR (ECONOMIST)| SEGOV/COPIFOR - CITY HALL OF FORTALEZAFortaleza, Brazil
Jun 21, 2024 7:28 AM
Replying to Sergio Luis Conte
...
There are framewoks to create prompt. This is part of the Prompt Desing discipline. Those that gave me and the initiatives where I was included are:R-T-F (Role-Task-Format), T-A-G (Task, action, goal), B-A-B (Before, after, bridge), C-A-R-E (context, action, result, example), R-I-S-E (role, input, steps, expectations).
Thanks for sharing, Sergio. You are giving us the opportunity of knowing more formulas for create effective prompts and, consequently, improving our learning and results with GenAI. Saving Changes...
Ekhlas mohamedSoftware Development Lead| RIDA GROUPKhartoum, NR, Sudan
Absolutely—refining a prompt can completely transform the output. I’ve had moments where a vague prompt gave me a generic or off-topic response, but adding just a bit more context or rewording it made the answer way more accurate and useful. It’s like giving better directions: the clearer you are about what you want, the more likely the AI will “get it.” I’ve even seen situations where simply specifying the tone or format (like “write this as a professional email” or “explain it like I’m five”) made a huge difference in quality. Saving Changes...
Anonymous
Refining a prompt can dramatically change the quality of AI-generated responses—it’s like adjusting the lens on a camera to get a sharper, more precise image. Here’s how it can make a huge difference:
1. Clarity and Specificity Unlock Depth
2. Structured Formatting Improves Readability
3. Prompt Engineering Enhances Reasoning
4. Context Ensures Relevance
5. Adaptive Refinement for Iteration Saving Changes...
Refining a prompt can dramatically improve the quality, accuracy, and usefulness of a GenAI output. In my experience, even small changes in wording or structure can shift results from vague or irrelevant to clear, actionable, and well-aligned with your goals. Here are a few specific examples of how refining prompts has made a big difference:
Example 1: From Vague to Specific
Original prompt:
“Explain cloud computing.”
Refined prompt:
“Explain cloud computing to a group of non-technical business stakeholders in under 200 words, using simple language and one real-world analogy.”
Result: The refined version produced a clear, concise, and audience-appropriate explanation, whereas the original was generic and overly technical.
Example 2: Sharpening Goals
Original prompt:
“Give me a project plan.”
Refined prompt:
“Create a 6-month project plan for migrating a legacy banking system to the cloud, with milestones, risks, and resource needs.”
Result: The refined prompt delivered a detailed, relevant plan. The original yielded a basic, non-contextual template.
Example 3: Guiding Format and Tone
Original prompt:
“Summarize this report.”
Refined prompt:
“Summarize this report for an executive briefing, highlighting KPIs, financial risks, and action items in bullet points.”
Result: The second version was far more useful for stakeholder communication.
Iteration helps me a lot
At the moment I have limited experience with this tool but I immediately realized how you can easily get to the results you need talking to AI as you do with a real person. You iterate and give more details in the direction you need. You specify where to dig more and what to skip. I'm really willing in practicing more. Saving Changes...
There are framewoks to create prompt. This is part of the Prompt Desing discipline. Those that gave me and the initiatives where I was included are:R-T-F (Role-Task-Format), T-A-G (Task, action, goal), B-A-B (Before, after, bridge), C-A-R-E (context, action, result, example), R-I-S-E (role, input, steps, expectations).
Perhaps my approach is too simple but I have found that having a "conversation" with the AI tool actually improves the results and then I apply the CREATE approach if further refinement is needed. I sometimes feel odd chatting with AI in this manner but so far, it has resulted in the outputs I find most useful. Saving Changes...
The output/response is succinct, clear and concise when the prompt is well crafted or structured. Saving Changes...
Damian WeitherspoonProject Engineer| Rheem ManufacturingMontgomery, Al, United States
Jun 21, 2024 9:36 AM
Replying to Eduard Hernandez
...
Increasing specifity and more context leads to more accurate and refined results/output. I am not familiar with the diverse frameworks provided by Sergio Luis Conte; speaking to GenAI engine as I would speak to another human (thus, providing context and sufficient level of detail) provides great outputs.
I agree. By adding more and more specificity, we can home in on more accurate results in an iterative way. Saving Changes...
PMI Team Member
