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?
Amy HiltunenProject Manager| National Association of State Workforce Agencies (NASWA)Mi, United States
First, stopping to think about what I'm really trying to ask and what I what has helped me, instead of just jumping in and typing. The prompt formulas have been really helpful as well. Saving Changes...
Angelo SalazarFounder & CEO| Avanza MethodsLos Angeles, 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).
This fantastic! I haven't heard of a couple of these, thank you.
Saving Changes...
Angelo SalazarFounder & CEO| Avanza MethodsLos Angeles, CA, United States
Great thread. As discussed throughout, there are many frameworks we can use for different purposes. But one of the things I find works regardless of the framework or topic, is to add boundaries to the prompt. To the extend that you can, try to think of things that should not be in the reponse given by the GenAI model. It's like the "Out of Scope" section of the Project Charter. Doing so focuses the AI and generally improves the quality of the output the AI generates. Saving Changes...
Badri N SrinivasanAVP and Lead - Agile Center| Societe Generale Global Solutions CenterBangalore, Karnataka, India
Refining a prompt has drastically changed the output quality for me many times. Initially, the output may not be so good. However, by refining the prompt, the output quality drastically improves and new insights are gained. Saving Changes...
Refining the prompt by chaining the responses, and by breaking down the problem in its parts, has given me much better results. Giving examples also helps the AI understand what kind of answer you are expecting.
Refining prompts in GenAI would help to improve output quality, of course clear and precise instructions is important for AI generate more accuracy output. Saving Changes...
Ramesh AnandTechnical Project Manager| PhilipsBangalore, Tn, India
In my experience with GenAI, using BAB - Before, After, Bridge Technic with iterations has refined a prompt drastically that changed the output quality more than expected! Saving Changes...
Robert David WatkinsChief Relationship Officer| Global Internet EntrepreneurCumberland, Md, United States
I have combined iterative measures with an interviewing technique wherein the AI interviews me with a series of "fleshing out" questions and clarifications and iterates to produce a final prompt, written by the AI, and aligned to a general initial objective prompt. Learning to interact with the AI such that a refined AI-produced prompt is generated and the back and forth "interview" has facilitated its production to include examples where the AI has requested them and all the CREATE framework elements. Saving Changes...
I have not had try the gen Ai support, but as I have learned this will help the PM a lot, then I am starting using it rigth now. When i have some ifnormation to dhare I will ocme back Saving Changes...
Fine-tuning of the prompts in GenAI is as delicate as carving out a statue; any changes made to the design altogether change the output. The prompt has to be well defined, this means that it is textured with detail and particularity to lead the model clearly to the intended destination. For instance, an ill-defined prompt such as ‘write a story’ will likely result in a generic story but a well-defined one like ‘write a dystopian sci-fi short story about a robot uprising in a futuristic metropolis’ will prompt the model to create better content. In addition, other parameters could be included in the recipe, such as a certain set of keywords, particular phrases or certain styles have to be included. When especially fine-tuned, the prompts allow accessing the full potential of GenAI, guaranteeing that it produces content suitable for achieving the business objectives, adequately credible, and appealing to the target audience. Saving Changes...
PMI Team Member
