Project Management

Please login or join to subscribe to this thread

When using AI systems, what are some best practices for ensuring the results you receive are accurate, relevant, and aligned with your original goals?

linkedin twitter facebook   Artificial Intelligence  
avatar
Sarah Philbrick
PMI Team Member
Director, Learning Design & Development| PMI Asheville, 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?
Sort By:
< 1 ... 110 111 112 113 114 115 116 117 118 119 120 ... 208 >
avatar
Anonymous
This may involve including the contextual factors in a detailed way and selecting an AI tool/system that will provide relevant and accurate details.

It should also include revalsoby a human being with relevant knowledge and expertise.
avatar
Anonymous
Don't apply the results directly. First prototype and validate.
avatar
James Moll CIO| Pretty Pickins LLC Douglassville, Pa, United States
Experiment with the various patterns, and then have a test or sample pool of inputs and expected outputs with which to evaluate the various prompt patterns.
avatar
Hasan Nazmul PMO Lead| Nitto Digital Dhaka, Bangladesh
For me, getting accurate and relevant results from AI starts with being crystal clear on what I’m trying to achieve. If I can’t explain my goal in plain language, the AI won’t magically guess it for me.
I make sure to give it solid context—background details, constraints, and examples—so it’s not filling in gaps with guesswork. Then I treat the process like iteration, not instant magic: I review the first output, tweak the prompt, and steer it closer to what I need.
I always cross-check important outputs against trusted sources or experts, and I keep some guardrails in place so it doesn’t drift off-topic or make up facts. And no matter how good the results look, I keep my own judgment in the loop—AI should speed me up, not make decisions for me.

Bottom line: AI works best when you guide it like a sharp, fast teammate—you set the direction, keep it honest, and make the final call.
avatar
Yves Olivier MOUANGUE ETOA Etampes, IDF, France
The challenge would be to go very specific and providing clear context with detailled information iteratively
avatar
Anonymous
Clearly define your goal, give specific and contextual prompts, and refine them as needed. Always verify results with reliable sources, watch for bias, and ensure outputs align with your original objectives.
avatar
Anonymous
Specificity
avatar
Maria Galvan Mexico
I treat AI as a helping hand, not as a replacement of my expertise. So I review all the answers an ensure context, judgment and nuance are preserved.
avatar
Ademola Kujore United Arab Emirates
Providing the specific context clearly and concisely, and keep iterating and fine-tuning the context given to GenAi
avatar
Carlos De Souza Brazil

I believe it's recommended to validate the information resulting from the research before publishing it, in addition to providing feedback to the model so it can refine future queries related to the topic.



• Ask the model for the sources of the information to verify its veracity and currency;



• Check for sensitive, copyrighted, or proprietary information;



• Check that the information and its format are appropriate for the target audience;
< 1 ... 110 111 112 113 114 115 116 117 118 119 120 ... 208 >

Please login or join to reply

Content ID:
ADVERTISEMENTS

"The industrial revolution was neither industrial nor a revolution - discuss"

- Linda Richman

ADVERTISEMENT

Sponsors







Newsletters

Sep 9: Becoming the 'Top Gun' of Your Project
Sep 2: Busy is Not a Symptom of Ambition
Aug 26: Stop Writing Status Reports

See all newsletters