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?
Michelle JubenvillProgram Administrator| University of ManitobaDauphin, Manitoba, Canada
Just beginning to work with AI; great session! Saving Changes...
Anonymous
Provide context, properly number or order the prompt if you are requesting multiple steps, provide templates or examples if you are requesting a specific format of the response, and iterate through response/refinement of question, Saving Changes...
Dhanwada RaoEnterprise Data Program Manager| ComericaRochester, Mi, United States
Articulating the expected outcome and refining the tasks for AI to perform shall provide best response. Illustrating a scenario (protecting sensitive info) to use as an example further hones the AI to provide better response Saving Changes...
folusho obePM Consultant| Phamedics Consulting, LLCZebulon, NC, United States
All the responses above are good responses and I would just add to be specific and adjust/refine as you go to ensure the desired output. Saving Changes...
Anonymous
As in all outputs you need to restructure to what is natural to you, check double check and ensure what is produced is actually answering the issue and not just filling space. Saving Changes...
It’s not just about asking questions but knowing how to frame them. To get accurate results, you need to define the right persona, craft detailed prompts, and provide examples of the outcome you’re aiming for. Add references where relevant and evaluate the responses based on what you need. It’s all about clear communication and being specific with your input to get the best out of the system Saving Changes...
You need to be specific on the tasks to be performed, give context to the AI model based on reliable information and give examples that aligned to your company´s objectives. Saving Changes...
David Andrés MoraTechnology Consulting | Data & Business Analyst | AI Enthusiasts| n/a
Hi everyone,
Based on my experience, to ensure that AI-generated results are appropriate, the following points should be considered:
(1) Validate the sources of information to ensure credibility and accuracy.
(2) Always define the appropriate context to guide the AI in generating relevant outputs.
(3) Review and refine the results by iterating, adjusting, and validating the information to ensure alignment with the objectives.
(4) Assess the criticality of the deliverable and have an initial idea along with possible alternatives. This can be achieved by consulting with colleagues who are experts in the field or by referencing formal and reliable information sources. Saving Changes...
The extent to which I trust AI results depends on my own knowledge. If I have no prior understanding of the topic, I am more vigilant, doing my best to validate the results and cross-check them with other references. In areas where I have expertise, I adhere to established principles and ensure the outputs align with my understanding
Saving Changes...
Ma. Carmen Lontoc Testa, PMP, PCCCertified Holistic Life, Career and Executive Coach| Independent ConsultantBonifacio Global City,Taguig, Philippines
Determine the specific task you want AI to work on, use CREATE to write a prompt that more clearly instructs AI what you want it to do/produce, and continue to refine the prompt as needed until you get the results you want and need. Practice the different approaches in creating prompts to see what works best for the subject/project. Saving Changes...
PMI Team Member
