Project Management

Please login or join to subscribe to this thread

In your experience with GenAI, how has refining a prompt drastically changed the output quality?

linkedin twitter facebook   Artificial Intelligence  
avatar
Sarah Philbrick
PMI Team Member
Director, Learning Design & Development| PMI Asheville, 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?

Sort By:
< 1 ... 150 151 152 153 154 155 >
avatar
Jose Antonio Morales Consultor, Analista, Evaluador Tecnico| Consultoria y Supervision de Proyectos SAC Lima, Lim, Peru
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.
speaking to GenAI engine as I would speak to another human, is not the same.
avatar
Carter Gray Pune, India

Adding clear context, constraints, and examples can drastically improve the output. Instead of a vague prompt, specify the audience, goal, tone. Even a few extra details can make the response much more relevant and accurate.

avatar
Aleja Seabron United States
Sometimes the answer is refining the prompt, sometimes it's moving to a different platform.
avatar
Andrés Salas Reyes Lider de Riesgos| Antofagasta Minerals Santiago, RM, Chile
In my experience, the quality of the output improves significantly when the prompt is structured with greater precision. What works best for me is defining the AI’s role, the context, the objective, the required activities, the expected result, and the desired output format.
I have also found that an iterative approach produces better results: working step by step, reviewing each response, and providing clear feedback when the AI is either aligned with or moving away from the expected outcome.
In summary, a good prompt is not just a well-written question; it is a clear specification of the work, combined with continuous refinement through feedback.
avatar
Anonymous

Always be specific

avatar
Tanja Josipovic Program Manager| Paycor Belgrade, Serbia
When creating project updates, I used to ask AI to "summarize project status." After refining the prompt to include the target audience, key focus areas (risks, milestones, dependencies), and preferred format, the output became significantly more actionable and executive-ready. Prompt refinement has reduced editing time and improved the usefulness of AI-generated content
avatar
Dennis Burns Special Projects Manager| ABM Building & Industry Long Beach, Ca, United States
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).

Refining prompts helps me extensively. There are times when I find that the AI is feeding back incomplete information, so I thoughtfully provide more information.

avatar
Maged Fowzy Project Manager| United Arab Aluminium Ajman, Ajman, United Arab Emirates

The more you refine, the more you get accurate answers.

avatar
Bhupesh Garapati Technology Project Manager| CrimsonLogic Pte Ltd Singapore, Singapore
In my experience with GenAI, I have seen a significant difference when I refine a prompt by adding the right context.
I may start with a simple question, but the output improves considerably when I clearly explain the objective, background, expected role of the AI, constraints, and desired outcome. I also use follow-up prompts to challenge assumptions, ask for deeper analysis, or simplify the response for a specific audience.
For example, instead of simply asking AI to identify project risks, providing the project context, dependencies, timelines, known constraints, and asking it to avoid assumptions produces a much more relevant risk assessment.
One thing I have learned is that prompting is not always a one-time activity. I see it more as an iterative conversation provide context, review the response, refine the prompt, and validate the final output.

The quality of the context we provide often determines the quality of the outcome we receive.
< 1 ... 150 151 152 153 154 155 >

Please login or join to reply

Content ID:
ADVERTISEMENTS

"Thousands of candles can be lighted from a single candle, and the life of the candle will not be shortened. Happiness never decreases by being shared."

- Buddha

ADVERTISEMENT

Sponsors