Director, Learning Design & Development| PMIAsheville, NC, United States
Validating and checking outputs is critical when working with AI systems like Generative AI. Such validation approaches may include establishing clear criteria, implementing strong testing protocols, and continuous refinement.
In your experience with AI, what are some best practices for ensuring the results you receive are accurate, relevant, and aligned with your original goals?
A good question. Clearly define your goal and provide specific content. Use precise, detailed prompts with relevant constraints. Verify important facts using reliable resources. Review and refine the output through follow-up prompts and check the response that it aligns with your original objective.
Saving Changes...
PO-AN HSIEHProject Manager| Timing International Co.Taipei, Neihu, Taiwan
When using AI systems, a good practice is to start with clear and specific instructions so the system understands your goals, then critically evaluate the output by cross-checking with reliable sources. Keep refining your prompts if the first response isn’t aligned, and always verify that the information connects back to your original objectives. This way, you ensure the results are accurate, relevant, and truly supportive of your intended purpose.
Providing the specific context in clear and consise way is essential.
I completely agree. Clear and concise context is the foundation for high-quality AI outputs. I would add that validation shouldn't stop there. It's also important to define clear success criteria, verify AI-generated information against reliable sources, and iterate through follow-up prompts to refine the result. In my experience as a project manager, the best outcomes come from treating AI as a collaborative assistant whose outputs require critical thinking, domain expertise, and human judgment before being used in decision-making.
Saving Changes...
Anonymous
CREATE is the best method that has helped me in refining Prompts
Saving Changes...
OLUWAROTIMI OLUWATOYINBOProject Engineer| Newfoundland and Labrador HydroSt. John's, Newfoundland, Canada
h1/h1h1/h1h1/h1h1/h1h1Use clear goals, give context, and define constraints so the AI understands your intent. Break complex tasks into smaller prompts, verify outputs against trusted sources., and iterate to refine accuracy. Ask the AI to show reasoning, test alternative prompts, and continuously align results with your original objectives to avoid drift. /h1 Saving Changes...
Liam SoperProject Manager HPM| PhilipsBrighton, Ma, United States
So far from this module I have learned:
-it’s imperative to chunk your desired output and questions into smaller more manageable pieces then gradually build from there -be very specific in your ask: explain specialized industry terms or company specific info (without confidential information of course -be precise and clear about the expected output and what hat you want the IA to wear -be specific with the outcomes you are expecting -be sure you provide the outcomes you are expecting -test a couple of different ways of asking and ask IA for advice on a better way of asking the question Saving Changes...
Liam SoperProject Manager HPM| PhilipsBrighton, Ma, United States
From this course I have learned: -it’s imperative to chunk your desired output and questions into smaller more manageable pieces then gradually build from there -be very specific in your ask: explain specialized industry terms or company specific info (without confidential information of course -be precise and clear about the expected output and what hat you want the IA to wear -be specific with the outcomes you are expecting -be sure you provide the outcomes you are expecting -test a couple of different ways of asking and ask IA for advice on a better way of asking the question Saving Changes...
By providing examples of the desired product, being very specific on expected result, and listing details down to the font and color palette of a document.
PMI Team Member
