Project Management Central
|
Please login or join to subscribe to this thread
|
|||
|
|||
Christiaan van den Berg
Project Management Consultant| Project & Business Consultancy Pty Limited
Burlington, Ontario, Canada
Chain of thought is a good approach and ReAct. References should be consulted and personal judgement on own knowledge base and experience.
This remind me of early days with design software where you also needed to assess practicalities and realism of the design solution. You shall have to do the same with AI responses "Visualize the Reality". 
- Being clear on the context that we're about to ask (type of company, # of resources, what's the objective).
- Also being very specific on what are the outputs that we're expecting from GenAI. - I use to use an "altered" template as a document to guide GenAI to provide responses as I expected. - Make sure that none of conflicting information appears on the prompts - If AI does not provide expected answers, reread the questions, refine the questions and adjust accordingly - Last, check with team members and ask feedback from them and stakeholders for verifications. 
Some practices for ensuring the results receive while using AI are accurate, relevant, and aligned with the original goals include the following:
- First is to ensure that the questions asked AI are of quality and neither too brief or too wordy. - The context of the question or request must be properly explained - The goal that needs to be achieved must be explicit and clear The task expected of the AI must be clear - The type of response and the format must be very clear and set early 
Snehal Jibkate
Pune, India
Two main steps are 1) use a LLM with ethical and secure way as per company policies, to avoid legal issues, 2) check every output and change according to your need.
Be clear, precise and provide context
Mahmud Momoh
Lab Tech| PepsiCo
When engaging with AI systems, it is crucial to articulate your goals clearly and supply specific inputs that include adequate context to achieve precise and pertinent outcomes. I think it is very important to begin with general inquiries and progressively narrow them down through iterative refinement, employ follow-up questions to enhance clarity in the responses received. This can ensure a better use the systems, and the information receive aligns with the needs and goals of the project.
Then, critically assess the information by cross-referencing it with credible sources while remaining aware of possible biases. Point out inaccuracies or errors to help refine the system's responses, because AI's responses are based on the data it was trained on, which may not cover all scenarios or the most recent events. 
Use CREATE formula so as the guide, and make sure the data source is accurate, timely and clean.
Mevwerosuoghene Egbegbadia
Elkridge, Md, USA
Using the CREATE formular helps in giving accurate results.
Refine your prompts and provide specific questions. Iterate your prompts and get the specific and detailed answers.
JUAN CAMILO HERNANDEZ MAHECHA
PM I| Federacion Nacional de Cultivadores de Palma (FEDEPALMA) - Colombia
Bogota D.C., Cundinamarca, Colombia
Best practices for ensuring optimal results using artificial intelligence should focus on providing clear guidelines with sufficient information and context regarding the desired outcomes. Additionally, including examples and supplementary information in the prompts allows AI to respond in a more specialized manner.
|
|||
PMI Team Member
Please login or join to reply
ADVERTISEMENTS
|
"Every child is an artist. The problem is how to remain an artist once he grows up." - Pablo Picasso |
