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?
In my experience, refining a prompt has yielded results I wanted but also in some cases not made a difference. My question is how much time should be spent in refining a prompt, and when should i recognize that even with refinement, AI is not doing a great job. Given that my work is restricted in what I can share with an AI platform, sometimes refining is useless because AI just does not have enough information. Saving Changes...
Anonymous
Using the role significantly improved outputs. Saving Changes...
Typically, I use a prompt-chaining model to refine AI outputs, however my experience has been largely unstructured. This module has helped to provide frameworks, processes, and tools, to make me a better prompt engineer. In the past, my methodology would be to ask an initial question and then provide details, context, or omissions in future prompts. While this often led to a suitable answer, it has not been the most efficient process and has no parameters for documents successes vs lessons learned, often leading to the same issues/roadblocks. Understanding how to develop a more detailed initial reponse (using the CREATE formula) and then how to further refine to get the desired outcome will be fruitful in my future experiences with GenAI. Saving Changes...
Hemangini ShahSenior Business Intelligence Analyst| University of MichiganAnn Arbor, Mi, United States
These are incredibly helpful tips, not just for PMs but for anyone using AI in prompt engineering. Being clear and specific enables the AI to deliver more valuable answers, and it continues to improve as you ask more questions over time. Saving Changes...
Lenin González GarcilazoMobile Digital Solutions Director| Orion Innovation LATAMMéxico, Nuevo Leon, Mexico
when I started prompting I didn't use to get what I want, because I didn't know how to do it well, so the first idea was providing context, then I provided the role and the way I expect the output. Now It's good to know that by using formulas like RTF/CREATE and patterns for iterating and refining will provide accuracy, relevance from AI Models Saving Changes...
Mimi CuffProject Manager| Conference of State Bank SupervisorsWashington, Dc, United States
I haven't used GenAI yet. I have used chat GPT without the insight of the course that I am currently taking. When I am done the course I will play around with using the formulas provided so that I can further refine my prompts for better results. Saving Changes...
Franklin CurayProject Manager| EBTELQuito, Pichincha, Ecuador
I asked the AI to generate a risk matrix with heat map, however, the result was not what I needed. So I gave him examples and specified what I wanted and he gave me the results I expected. Saving Changes...
Tessa BergemannProject Management| Enterprise VisionsMinneapolis, MN, United States
It has completely changed the output quality. It is a way to double down into the request initially asked, to gain further detailed information. I enjoy it because you can separate pieces out, and the main request stays in the chat, so you don't have to reiterate yourself again. It saves a lot of time as well not having to retype everything out again from the initial prompt. Saving Changes...
Michael ZambonPrinciple Digital Product Manager| American Water CompanyWoolwich Township, Nj, United States
In my experience with GenAI, refining a prompt has drastically changed the output quality. For example, when I started with a vague prompt, the responses were often too broad or not quite on target. However, by adding more context, specifying the desired format, and breaking down the request into smaller parts, the AI's responses became much more accurate and relevant. Iterative refinement, where I adjusted the prompt based on initial outputs, also helped a lot. This process allowed me to guide the AI more effectively, resulting in outputs that were closely aligned with my goals and expectations. Saving Changes...
Anonymous
This has made resolution and feedback faster and easily achievable. Saving Changes...
"Technology is a gift of God. After the gift of life it is perhaps the greatest of God's gifts. It is the mother of civilizations, of arts and of sciences."
PMI Team Member
