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 ... 58 59 60 61 62 63 64 65 66 67 68 ... 191 >
avatar
Anonymous
A best practice would be to validate your responses by checking the source data and performing experiments
avatar
Sharice Davis Chicago, Il, United States
Best practices to use with AI systems to ensure accurate and aligned outputs include utilizing known formulas, providing relevant context, and iterative prompt refining. Formulas such as RTF or CREATE are simple, easy to use methods to prevent vague or unstructured outputs from the AI system. Further, sharing additional context or details helps the AI system produce outputs that are specific to the user's problem or challenge. Finally, iterative prompt refining will hone in on generic or misaligned information and results in a high quality response with actionable insights for the user.
avatar
Jacqueline Langwell Huntsville, Texas, United States
I always reference back to the name of the product. AI is a LLM. That means it's programming. I need to tell it what to do. It can pull general information from a database, but to apply that knowledge, it needs to know exactly what I need. Not vague, but explicit. I try to apply the old English report writing from high school...who? (who is asking and who is the audience receiving), what? (what exactly are you asking?), when? (the timeline is significant to the response), why? (why do you need this information?) and how? (how will this information be used?). Just starting with these basic questions help me to form a prompt.
avatar
Franklin Curay Project Manager| EBTEL Quito, Pichincha, Ecuador
Currently people are using AI to do their things faster, for example documentation, however, we must validate the responses obtained because sometimes they generate things outside of what we expect.
avatar
Tessa Bergemann Project Management| Enterprise Visions Minneapolis, MN, United States
The biggest take away I have learned is about how precise the prompt is - you can separate it out and iterate responses along the way. If you request too much all at once, like discussed in the video, some things can be missed. It is increasingly important to break it down as well, making sure you aren't asking vague questions about basic industries. Trying your best to hone in on the specific industry and requests helps confirm accuracy. I also do like the recite your sources prompt, because sometimes those can even turn up to be "hallucinations" and incorrect. Always make sure to double check what the AI has stated, because you know what you are requesting more than it does.
avatar
MOHAMMED NILE C, Bangladesh

There is no simple answer to this. I feel that it is a learning process and the basics must be covered. Using RTF or CREATE formulas would be the starting points. Also giving specific details about the contexts and parameters to be considered would be very important for the desired result.

avatar
Laura Lazzerini
Community Champion
Head of International Project Management Office| Deutsche Telekom Praha, Czechia
Make the prompt as much precise as possible, by indicating, the role, the situation, the kind of output that you wish to receive
avatar
Michael Zambon Principle Digital Product Manager| American Water Company Woolwich Township, Nj, United States

Be Clear and Specific: Clearly define what you want the AI to accomplish. Avoid vague language and be as specific as possible about the task or question.



Provide Context: Include relevant background information and context to help the AI understand the nuances of your request. This can include details about the subject, the desired format, and any specific constraints.



Use Structured Prompts: Break down complex requests into smaller, manageable parts. Structured prompts can guide the AI more effectively and yield better results.



Set Expectations: Clearly state your expectations for the output, including the level of detail, tone, and style. This helps the AI tailor its response to meet your needs.



Iterate and Refine: Don’t hesitate to refine your prompts based on the AI’s initial responses. Iterative adjustments can help you hone in on the most effective prompts.



Ask Follow-Up Questions: Use follow-up questions to clarify or expand on the AI’s responses. This can help you get more accurate and relevant information.



Incorporate Examples: Providing examples of the desired output can help the AI understand what you’re looking for and produce more aligned results.



Monitor and Adjust: Continuously monitor the AI’s performance and make adjustments to your prompts as needed. This ongoing process helps maintain accuracy and relevance.
avatar
Anonymous
Ensure that the request made is short and concise.
avatar
Jose Manuel Marrero de los Santos PMP| INTEGRADOC BPM Montevideo, Montevideo, Uruguay
Break down complex problems.
Iterate by providing feedback.
Validate sources.
< 1 ... 58 59 60 61 62 63 64 65 66 67 68 ... 191 >

Please login or join to reply

Content ID:
ADVERTISEMENTS

"To stimulate creativity, one must develop the childlike inclination for play and the childlike desire for recognition."

- Albert Einstein

ADVERTISEMENT

Sponsors







Vendor Events

See all Vendor Events

Newsletters

Jun 3: The Career Problem Facing New PMs
May 20: Your PM Career as a Project Life Cycle
May 13: The Big Lie After Layoffs

See all newsletters