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?
Markus SchuldtSenior Project Manager| ITERGO GmbHAhrensburg, Germany
As always you have to check the result. In particular, whether the result is understandable, comprehensible, logical and meets the requirements. If this is not the case, then I revise the prompt in the next iteration. Personal learning and development is essential here. Saving Changes...
Eyad HasanProject Director| IDEMIADubai, United Arab Emirates
Some ideas;
Provide the context
Clear request to the tool
add format of the output
it would be great if you add some examples of what you need
Inform the AI tool what criteria you are looking for to ensure precise and accurate results
One will need to incorporate patterns to fine tune the prompts and provide additional information that may be required tailored to your expectation and format. In addition be mindful of the ethics and standards Saving Changes...
Murthy SNHead - Service Excellence, Transformation| Ernst & YoungBangalore, Karnataka, India
While interacting with AI systems, I would pay a lot of attention on considering the appropriateness of asking some question to the system. Every question fed into the AI tool has to fetch a factual response ' else it is utterly useless. Few things I will look for is : (a) how much of latest information the tool has (b) does the tool have any bias ? (c ) does it prompt me if i share sensitive / confidential data to be on alert (d) does it give me references of similar scenarios or prompts used by others and refine responses (e) Human in the loop is critical ; I would be very keen to know how often the tool is updated with new information and what are the sources it taps into Saving Changes...
Preparing a well-formed prompt is essential. For simple inquiries, use the RTF model, while for complex and long-term tasks, apply the CREATE model. Validation is crucial. Establish a list of factors to be considered throughout the process. Finally, ensure end-user input is incorporated to effectively tailor the result to the specific mission or objective. Saving Changes...
Gaurav TannaIntech Creative Services Pvt. Ltd.Gandhinagar, Gujarat, India., Gujarat, India
While interactive with any AI tool e.g. chatbot, practice to ask Iterative Prompt through CREATE formula, which will give you more actionable output that are more aligned to your expected goals. Saving Changes...
Gaurav TannaIntech Creative Services Pvt. Ltd.Gandhinagar, Gujarat, India., Gujarat, India
While interactive with any AI tool e.g. chatbot, practice to ask Iterative Prompt through CREATE formula, which will give you more actionable output that are more aligned to your expected goals. Saving Changes...
results you receive are accurate, relevant, and aligned with your original goals?
A very good question that needs to be considered when dealing with LLM chatbots. In my opinion, for us to get the best out of the AI that is most accurate, relevant, and aligned with our intended request, we need to consider the following:
1. Adopt a clear framework/ structure when starting the chat (ex: RTF, CLEAR). This enables us to better frame the issue at hand, and help the AI to better visualize the context and setup.
2. Provide well-structured prompts that best fit your needs. This process will usually be iterative and will need to refined and fine-tuned in the chat context to converge to the desired goal.
By following these two main steps, we make sure that enough information is provided to the AI that enables it to give more straight-to-the-point responses that best serve our goals. Saving Changes...
If a woman has to choose between catching a fly ball and saving an infant's life, she will choose to save the infant's life without even considering if there is a man on base.
PMI Team Member
