Project Management Central
|
Please login or join to subscribe to this thread
|
|||
|
|||
Using the formulas and iteration along with feedback cheek is a best practice
Best practices include interactiving approach / ReAct for sure as we keep validating the output received to refine them to add optimal value for your requirement context. Let's also ensure to utilize opportunities if any from the halluzinated responses to elaborate the prompt further to get additional value to the context.
Karthick Nivas Ramdoss
Mumbai, Maharashtra, India
Few things that have helped me: a) Given enough context: define persona, examples, use-case, upload prior documents b) Clearly mention how the response will be measured/tested by you c) Iterate with the LLM or restart with a clearer question d) Try a different LLM
Aida Ep SNENE
Tunis, 11, Tunisia
React, adjust, give examples if necessary and ask to refine
Define Clear Objectives: Start by clearly defining the goals and objectives you want to achieve with the AI system. This helps in setting the right parameters and expectations from the outset.
Data Quality and Relevance: Ensure that the data used to train and operate the AI system is of high quality, relevant, and representative of the problem you are trying to solve. Poor data quality can lead to inaccurate and biased results. Regular Monitoring and Evaluation: Continuously monitor the performance of the AI system and evaluate its outputs against your goals. This helps in identifying any deviations or inaccuracies early on. Transparency and Explainability: Use AI systems that provide transparency and explainability in their operations. Understanding how the AI makes decisions can help in assessing the accuracy and relevance of its results. Ethical Considerations: Pay attention to ethical considerations such as fairness, bias, and privacy. Ensure that the AI system does not perpetuate biases or misuse data, and that it operates within ethical guidelines. Human Oversight: Maintain a level of human oversight to validate and interpret the AI system's results. Human judgment is essential in ensuring that the AI's outputs are aligned with the original goals and context. Iterative Improvement: Treat the deployment of AI systems as an iterative process. Continuously gather feedback, make necessary adjustments, and improve the system to better align with your goals. 
Bisharah Saeed
Project Manager| ansrsource
In one word I would say experience matters. As my colleagues said above, there are no clear defined metrics for evaluation. From my experience, I have realized that unless you have sound knowledge of what you are asking from the AI, ensuring accuracy will be a challenge. I recently used it for data analysis. Since I was clear with my end goal, I was able to tweak things before I got what I wanted.
When using AI systems, it is essential to establish clear objectives and criteria for success upfront to ensure results align with your original goals. Implementing rigorous validation processes, such as testing the AI system with a diverse set of data, monitoring its performance over time, and regularly realigning or refining it as needed, can help maintain accuracy and relevance.
Eric Boyd
Sr. IT Project Manager - Voice Services| Vanderbilt University Medical Center
Smyrna, Tn, USA
When using AI systems, what are some best practices for ensuring the results you receive are accurate, relevant, and aligned with your original goals?
It's good practice to ask AI Systems to show or provide a reference to your material. Perform a prompt to validate your response. Also, leverage your own experience or the experience of a subject matter expert. This will also make sure the response or accurate and relevant. Then it is always a good practice to review your original task or problem you are solving. If it is an iterative prompting taking place then review the response with the latest question. Don't be so quick to move forward with reviewing and editing the responses from AI. 
More and more practice required, also the data has to be well tested and trained for proper validation. Explaing the context is very important and expected outcome specification in detail make your job easy.
Keep your prompts specific and focus on mentioning the tone. Give precise and clear instructions as if telling a child what to do.
|
|||
PMI Team Member
Please login or join to reply
|
"Life begins at 40, but often so does arthritis and the habit of telling the same story three times to the same person." - Sam Levenson |
