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?
Kiruba SankarArea Coordinator Refinery Projects| Engineers India LimitedLagos, Nigeria
I tried with complex prompts, Chatgpt provides less accurate results. But for Complex question broken down into smaller ones the result was awesome. Like Humans it does mistakes in certain prompts, & like humans upon rephrasing it slowly changes its answers without accepting the mistake :)..
After a series of probing questions, it eventually acknowledges its mistakes & do apologises for its mistake providing revised one.
It's a lot like teaching and guiding a growing child; we need to train the LLM according to our needs, and it can work wonders!" Saving Changes...
David JohnsonFunctional Manager| BlueCross BlueShield of AlabamaBirmingham, Al, United States
I have had some success with changing the prompts to get more detailed and better formatted results. Saving Changes...
John NjorogeProject Management| Equity Group HoldingsSabaki/ Athi river, Kenya
Yes, it has really changed both the quality of output and also the time took prompt engineering.
i would recommend this series of online learning to anyone who wants to make work easier and save time Saving Changes...
Anonymous
It helped narrowing the output to the concise and pragamtic version. Saving Changes...
HAMEED ALYAJORIAudit and Risk General Manager , Project Manager| Yemen Customs AuthoritySn, Yemen
Jun 23, 2024 8:30 AM
Replying to Sergio Luis Conte
...
You are welcome Winston C Ikekeonwu PMP . I spend my time here to learn from all people comments then my gratitude is with you. Just to add to your comment about input data you are totally wirte. It is a key in all AI and mainly in generative AI. That´s the point from long long time ago and as you know now is "converted" a new discipline call Data <something>
Refining a prompt can drastically improve output quality by providing clearer context, specific instructions, and desired formats. A vague prompt often leads to generic or irrelevant responses, while a well-structured one ensures precision and relevance. Adding examples or constraints helps guide the AI toward the intended outcome. Iterative refinement allows users to identify and address ambiguities. Ultimately, a refined prompt aligns the AI's response more closely with user expectations. Saving Changes...
Refining a prompt by being more specific on the request (provide details and explain any acronyms or technical words) and providing examples and formats for the response leads to more accurate outputs. Saving Changes...
Christiana EtukProject Manager| Helium HealthUyo, Ak, Nigeria
Hello PMI Community, I am currently working on projects to deploy software to hospitals. However, before doing that, I must collate the hospital's billed services to configure the data into the software. The major risk is that most clients do not have the required device to review the shared template and fill out the necessary information, such as medication costs, consultation fees, laboratory investigation fees, etc. Using AI how can this major problem be solved? I need as much ideas and suggestions to help me pull through this hurdle. Saving Changes...
By refining the prompts using CREATE formula has significantly improved the output by making it specific, targeted and effective in my use of CoPilot. Saving Changes...
Swaminathan NProject Manager| ELGI EQUIPMENTS LIMITEDCoimbatore, TN, India
An wonderful and very informative insights of Generative AI. It is really a very useful tool if we understand how to extract qualitative and precise information. 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).
I've seen prompts go from vague requests to highly specific instructions, dramatically improving output. For example, a prompt like "write a short story" might yield generic results. But changing it to "write a 200-word science fiction story about a robot learning to feel emotions, set on Mars" provides the AI with much-needed context and constraints, leading to a far more focused and creative story. Adding keywords, specifying length, and defining genre are all ways to refine prompts for better results. Saving Changes...
PMI Team Member
