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?
Have tried to reduce the cache size as well as the size of the response output parameters (not in Copilot) to guage the output and performance. Varied response and still work in progress however. Saving Changes...
Puja MehraSr. Project Manager| ACI WorldwidePrinceton Junction, Nj, United States
I usually use the re-iterate method to get he output closer to my needs possibly by giving a few examples. 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).
These are great prompt frameworks. I was unfamiliar with a few of them. Thanks for sharing. Saving Changes...
Indeed, it is the first time that I am working on artificial intelligence and talking to GEN-AI as if I were talking to a human, and of course it is necessary to work on sound frameworks through which accurate answers can be obtained from the first time and the answers can be improved over and over again on the same topic to improve the outputs more and more. This has clearly improved my productivity. Saving Changes...
Anonymous
With GenAI, refining a prompt drastically can changed the output quality by tailoring formulas such as CREATE to do 2 or 3 task so AI can be more specific to what we want and align with our goals. That is the important aspect that the human loop has to be there to ensure we understand those goals and therefore refining the task accordingly Saving Changes...
Josh BehlAcademic Faculty Member| Minnesota State University MoorheadFargo, Nd, United States
oh my goodness, yes! Unfortunately, I have taken a trial and error approach when I first started, but as I continued to use, create, validate, and refine my prompts over the last couple of years, it has become considerably better. Having said that, I am learning a great deal in this course that I will take back and incorporate immediately into my prompt engineering. Saving Changes...
Caryn OgierDirector of Project Planning| The Quay Companies, LLCTx, United States
I used ChatGPT to help create a presentation on meeting management. It was generic at first but once I started defining the audience of the presentation, I was able to tailor the content accordingly to ensure it was relevant. Saving Changes...
Shayma IvankoProject Manager| High Tide Consulting Inc.Ajax, Ontario, Canada
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).
Other than RTF and CREATE, I was not aware of other frameworks for prompts listed above. I'm very interested in trying them out as I've experienced a comparable difference when generating responses with vague prompts vs a structured layout. The quality ends up being more specific to what is being asked, and correct. Saving Changes...
In my experience with GenAI, refining a prompt has often made a notable difference in the quality of the output. When I start with a vague or overly broad prompt, the results tend to be generic or off-target, but adding clarity and specificity unlocks much more relevant and insightful responses. I've found that including context, constraints, or examples in the prompt can transform an average result into something highly tailored and useful. Iterating on prompts also helps me guide the AI toward exactly what I need. It’s taught me that the time I invest in crafting a strong prompt directly translates into better outcomes.
Saving Changes...
Raffi JalakianGlob Program Associate Director | Amgen Inc.Boston. MA, United States
Adding more specifics to meet the desired outcome: I have asked AI to generate 2 versions of the risk analysis for 2 target audiences 1) Technical and detailed for the team of scientists and 2) a higher level for executive audience focusing on outcome and impact. Saving Changes...
PMI Team Member
