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?
Indeed, to obtain reliable results, the information provided must be the best. We are facing a situation that some of us have experienced several times before. Back in the 1970s, the "Slide Rule" was used, then calculators, portable calculators, spreadsheets, Power BI, specialized software, ERPs, data science, and other Artificial Intelligence options. The difference lies in the speed of change and the generation of results. The question we must ask ourselves is how many of us studied the scientific calculator manual, how many of us master Excel, how many of us know something about Power BI, how many of us master an ERP, and so on. This will not be the end of Project Management; it is a new opportunity to enhance our capabilities. But be careful, because these potentialities are appearing more and more frequently, multiplying capabilities several times over, and will address the most common needs in a very simple way, ones that will not require as much validation and verification. Saving Changes...
Bandar BahlulManagement| Logistics ManagementRiyadh, 1, Saudi Arabia
Best Practices When Using AI Systems:
Define your goal clearly from the start
Be specific about what you want to achieve—whether it's data analysis, report generation, or decision support.
Use precise and well-structured prompts (Prompt Engineering)
The clearer and more specific your instructions, the better the quality of the results.
Verify the results
Don’t rely blindly on AI outputs. Review them and, if possible, cross-check with other sources—especially for sensitive or complex decisions.
Ensure data and context are up to date
Make sure the system is using current and relevant data for your local or time-specific context.
Use AI as a support tool, not a replacement
AI enhances your capabilities, but it doesn’t replace critical thinking or human expertise.
Respect privacy and security
Avoid sharing sensitive or confidential information unless you’re sure the system is secure and authorized.
Saving Changes...
Gopal AsthanaService Coordinator and Incident Manager| KyndrylNoida, India
I’d say the key is treating AI like a partner rather than a magic box. Start by being clear about what you want - give context, set boundaries, and explain your goal in plain language. If the answer feels off, don’t hesitate to reframe or break your request into smaller steps. Cross-checking the response with your own knowledge or a trusted source also helps catch errors. And finally, keep in mind that AI isn’t perfect - so staying curious, iterative, and a bit skeptical is the best way to get results that actually serve your purpose. Saving Changes...
Anonymous
thank you for sharing. Saving Changes...
Anonymous
thank you for sharing. Saving Changes...
Anonymous
Very informative Saving Changes...
Ronald CairoSpecialist Engineer| Ministry of Housing, Construction and Sanitation of PeruElkridge, MD, United States
After defining the position of the person whose response is required, after specifying exactly what is being requested, the sources must be requested and one must always have an alternative solution; these will be compared with the answers provided. Saving Changes...
The key recipe to elicit accurate, relevant and aligned response from a genAI will always be the human in the loop. The knowledge of the subject can not be wholly subcontracted to a machine... The work of the AI is really to compress time and make you more productive. The knowledge of the subjectest is key ingredient in the drafting of a comprehensive prompt, refining the prompt and validating the output of a genAI.
...
1 reply by Ronald Cairo
Aug 25, 2025 6:22 AM
Ronald Cairo
...
Technology advances, continuous improvements never stop, the world has been changing since the beginning. However, I'm sure that more and better tools will continue to emerge; but they will never be able to replace the human being, the responsible person, the one who signs the document. They may approach and/or exceed pre-established limits, but they are not responsible for anything; the person is.
Saving Changes...
Ronald CairoSpecialist Engineer| Ministry of Housing, Construction and Sanitation of PeruElkridge, MD, United States
Aug 25, 2025 5:01 AM
Replying to leye adewoye
...
The key recipe to elicit accurate, relevant and aligned response from a genAI will always be the human in the loop. The knowledge of the subject can not be wholly subcontracted to a machine... The work of the AI is really to compress time and make you more productive. The knowledge of the subjectest is key ingredient in the drafting of a comprehensive prompt, refining the prompt and validating the output of a genAI.
Technology advances, continuous improvements never stop, the world has been changing since the beginning. However, I'm sure that more and better tools will continue to emerge; but they will never be able to replace the human being, the responsible person, the one who signs the document. They may approach and/or exceed pre-established limits, but they are not responsible for anything; the person is. Saving Changes...
Khalid IbrahimEndpoint System Support Analyst| Tangerine BankToronto, Canada
My best practices:
- Start with a clear goal + RTF prompt (Role, Task, Format).
- Ground the model in your data (docs/requirements) and cite sources when possible.
- Test in small chunks and compare outputs against acceptance criteria.
- Keep a human-in-the-loop: spot-check facts, numbers, and assumptions.
- Iterate (refine prompts, add examples), and keep a brief decision log.
- Mind privacy & bias: remove sensitive data and ask for uncertainty/limitations.
PMI Team Member
