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?
Rajakeerthy VijayakumarProject Manager| Shaw CommunicationsVancouver, British Columbia, Canada
Jun 23, 2024 8:31 AM
Replying to Sergio Luis Conte
...
Thanks @Abdur Rehman. You are welcome.
Refining the prompt from my experience, I get the output aligned to my expectation and goals. Saving Changes...
Robert LobdellProgram Manager| Chickasaw Nation IndustriesEnterprise, AL, United States
Refining a prompt can make a huge difference in the quality of AI output. A vague prompt might give you something generic, but when you add clarity, structure, and context—like defining the role, breaking down the task, or specifying the format—you get results that are sharper, more relevant, and aligned with your goals. It’s like giving better instructions to a smart assistant: the more precise you are, the more useful and tailored the response becomes. This kind of refinement turns AI from a basic tool into a real strategic asset. Saving Changes...
Being concise and specific helps the AI to give some valuable answers. It also learns with time as you ask further questions.
AI is powerful, but it’s not a mind reader. It thrives on clarity. When you give it the right context, constraints, and purpose, it responds like a well-trained assistant who gets you.
Imagine asking someone, “Can you help me get somewhere?” vs. “Can you help me get to the airport by 5 PM, avoiding traffic?”
The second one gives them a clear goal, context, and constraints — and they’ll give you a much better answer.
That’s exactly how AI works. The more specific and thoughtful your prompt, the more useful the response.
Sometimes the first result is “okay,” but not quite right. So, you tweak the prompt by adding limits, asking for comparisons and changing the perspective.
Each tweak brings you closer to the result you need. Saving Changes...
I think it is important to give the context and also to refine, asking for a different output in case that the first one is not completely suitable to our purpose or to the outcome that we were looking for. I think that consistency and preseverance in looking for the result, is crucial as well.
AI is powerful, but it’s not a mind reader. It thrives on clarity. When you give it the right context, constraints, and purpose, it responds like a well-trained assistant who gets you.
Imagine asking someone, “Can you help me get somewhere?” vs. “Can you help me get to the airport by 5 PM, avoiding traffic?”
The second one gives them a clear goal, context, and constraints — and they’ll give you a much better answer.
That’s exactly how AI works. The more specific and thoughtful your prompt, the more useful the response.
Sometimes the first result is “okay,” but not quite right. So, you tweak the prompt by adding limits, asking for comparisons and changing the perspective.
Each tweak brings you closer to the result you need. Saving Changes...
MARTIN URREGOProject Management| PUCPMiraflores, Peru
As this course emphasizes, effective AI work starts with a clear, self-contained, structured prompt aligned to the desired outcome. In my experience, iterative experimentation—prompt tuning, incremental refinement, and cross-model benchmarking—has been essential to consistently achieve reliable, high-quality results. Saving Changes...
Anonymous
when LLM is provided with more specific and refined prompts, it gives very straight forward answers reduces assumptions and aligns with our objective and lists actionable items Saving Changes...
Anonymous
when LLM is provided with more specific and refined prompts, it gives very straight forward answers reduces assumptions and aligns with our objective and lists actionable items Saving Changes...
Thomas TuranProduct Owner, Investments Data Quality| Thrivent FinancialMn, United States
I have found that refining my prompts with data if I have it, including my ultimate goal, asking it to consider the question in the arena of Investments, and specifying my desired tone of language have all helped me get much more useful content than when I started out asking very open ended questions. Saving Changes...
Thomas TrippanyProject Management| Element U.S. Space and DefenseNorth Hollywood, CA, United States
We should provide an initial general request which would provide a very broad and generic response. We could then further refine our prompts based on what is outputted by the LLM. If we aren't quite sure of how to further iterate our prompts, perhaps we could ask the LLM to ask us questions. Saving Changes...
In my professional use of AI for my company, I directly focus on role (ex. Project Manager, Project Manager, Financial professional) to provide the LLM context about what framework to use in the response. In addition, I also provide the LLM with background information. Finally, I find it useful to provide a template to frame its response. This approach gets me 80-90% of what I want in the first responses. From here, I will refine questions to finish off the remaining 10-20% of what is required. Saving Changes...
PMI Team Member
