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 ... 101 102 103 104 105 106 107 108 109 110 111 ... 147 >
avatar
Shannon Hrudka Exeter, Ontario, Canada
There has been times when I don't trust the response from LLM based on gut feeling (i assume because the prompt was not well engineered), so if you ask "Are you sure? How can you be so sure?" it will provide the background information used to provide the response. Also, a side benefit is that holding a machine accountable simply makes my laugh.
avatar
Grettel Fonseca Project Manager| Lionbridge

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



It has definitely help a lot to get better outcomes when you provide feedback to the LLM, in my experience when providing context, examples and some relevant data instead of giving general information has always gave me better results. I think working with AI is very iterative, and it usually requires some refining before my prompt get the expected results. It also help me when I asked how they get to that outcome, and it help to identify if the information is valid or not.

avatar
Dr. Afzal Haq Asif Al-Ahsa, 4, Saudi Arabia
Really learning a lot...
avatar
Dr. Afzal Haq Asif Al-Ahsa, 4, Saudi Arabia
Great job done
avatar
Jennifer Scatcherd Labor Efficiency Manager| Atlantic Health Morristown, Nj, United States
I love that I can keep a chat and then add to when I need additional information. It remember where we are in the convo and allows for refinement as needed.
avatar
Anonymous
Refining prompt with added context and examples allows the best opportunity for providing the richest and most compelling output.
avatar
Anne Stein Project Manager| Barisoft Consulting Media, Pa, United States
By learning to maximally interact with an imperfect (but powerful) system (LLM) with its own parameters, optimizers, and limitations - one can obtain competitive advantage in producing higher quality/targeted/value add/"aha insights"
avatar
Roderick Nunez Laguna, Philippines
Before when I do some Sales Analytics for one of our subsidiaries, I used to prompt "analyze this data set", over time, I am now including in the prompt the identification of seasonal patterns, data outliers, and asking for a maximum of 5 actionable insights to improve sales performance. !--StartFragment --The refined prompt leads to a better strategic recommendation. GenAI models basically treat your prompt like a set of instructions. The clearer and more specific you are, the more accurately they can produce the desired output.
avatar
Imran Rashid Project Manager| elGrocer Faisalabad, Pakistan
Applied "Prompt Chaining" to my simple and convensional prompt which was resulting in more generic results. Divided my complex task into different smaller sub tasks. Using CREATE formula, i left Character, Adjustment, Types of output and Evaluation parts as they were but changed Request and Examples. This approach produced more accurate results.
avatar
Ifeanyi Nwosu Lagos, La, Nigeria
Jun 21, 2024 10:50 AM
Replying to Kiron Bondale
...
Sarah -

This morning, one of my LinkedIn contacts complained that GenAI tools don't seem to have the ability to craft decent quality PMP practice exam questions. He had used the public version of ChatGPT. I decided to check the same with PMI Infinity and got better results - seven out of ten questions were acceptable.

My first prompt was "Generate ten different questions about project management which would be similar in style and level of difficulty to what is asked on the PMP exam"

It only gave me the questions but neglected to provide any answers. Realizing that this was likely it interpreting what I had asked for "as is", I then added: "Generate ten different questions about project management which would be similar in style and level of difficulty to what is asked on the PMP exam".

With that it was able to provided more useful output.

Kiron
Exactly my thoughts. When you are more specific with the outcomes you expect from the Gen AI tool in your prompt, you are more likely to get a good result.
< 1 ... 101 102 103 104 105 106 107 108 109 110 111 ... 147 >

Please login or join to reply

Content ID:
ADVERTISEMENTS

"We cling to our own point of view, as though everything depended on it. Yet our opinions have no permanence; like autumn and winter, they gradually pass away."

- ChuangTzu

ADVERTISEMENT

Sponsors