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?
Suggested Discussion Response: In my view, advanced prompting patterns can significantly enhance how project managers use AI to support decision-making and adapt to changing project conditions. Techniques such as ReAct or iterative prompting help ensure that AI outputs remain aligned with evolving objectives, constraints, and stakeholder expectations. By continuously refining prompts and incorporating feedback, project managers can use AI not only as a productivity tool, but also as a strategic support mechanism to improve governance, prioritization, and risk management throughout the project lifecycle. Saving Changes...
Usando a GenIA a comunicação com o cliente melhorou bastante, pois a precisão com os prompts dá mais insumos que geram credibilidade com dados.
Saving Changes...
Konstantin IvanovProject Manager, Co-Founder| Multiple Solutions F.Z.E.Dubai, DU, United Arab Emirates
I’ve seen a big difference when I take time to refine a prompt. Early on, I’d get generic or overly broad answers, but once I started clearly stating my goal, context, constraints, and what I actually wanted to use the output for, the results became much more relevant and usable. Breaking a single question into smaller, focused prompts has been especially effective in improving accuracy and practical value. Saving Changes...
The first answer is get is sometimes not exactly what I'm looking for. By refining the prompt and asking clarifying questions and well as by providing addition clarification and data, I'm able to get a much more accurate, comprehensive and useable answer.
Saving Changes...
Anonymous
No, in my experience with GenAI when I refined my prompt drastically, it normally would not change the output quality. I do not have much experience using GenAI, but when used it seemed to be confused or almost put out the same information (it could have been where it was "pigging back" off of my previous responses.
Saving Changes...
Joseph HannaVP of IT| Repipe SpecialistsDoha, None (International), Qatar
Based on what you are feeding the LLM model with, you will get the most tailored response. So start by using a clear and straightforward request, show and attach examples to fine-tune how the answer will be. Then start using one of the chaining methods to continue with your queries. If you find something is not right, always visit and teach and feed againt he LLM model until you get what you want. Remember, it all comes to the reading; do not depend on any answer without reading.
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).
PMI Team Member
