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?
To get relevant, accurate and usable content, it is all ways suggested to follow the prompting formulae CREATE, which will eliminate hallucinations and improper output.
Saving Changes...
Adane MollaConstruction Project Manager| Water Works Corporation (WWC)Kombolcha, Ethiopia
Creating an initial draft prompt. Submit to the ai model for refinement to correct the phrasing and wordings that are relevant to the problem context. In this way, the result that I get from the edited prompt by the ai itself gave me better or best results and outputs. Saving Changes...
Treat AI as a project team member. Give clear instructions by using the CREATE method for your prompts, including background context, what the end goal is, and clear examples when possible. You should also verify the output by comparing it to trusted sources. Always review the response to make sure it matches your original purpose and is useful for the task. Saving Changes...
Anonymous
To consistently get accurate, relevant, and project goal aligned results from AI system, you must provide clear and concise inputs and validate those inputs through an iterative process in which established workflows are used. You should start with clear and concise instructions, provide the proper context, specify desire format, and break complex problems into sequence of task/step. Use the iteration to refine.
Saving Changes...
Lawrence WaldenProject Manager| Chipotle Mexican GrillColumbus, Oh, United States
Be precise and clear.
Be sure you explain jargon or specialized terminology
Ask the AI to cite sources or list assumptions. Cross‑check important facts with trusted references. Request a confidence rating or ask: “What might be wrong here?” Saving Changes...
From my perspective, an effective practice for validating AI-generated data consists of developing a checklist with the project team. By relying on our team, we can streamline the verification process, thereby eliminating ambiguity. Nevertheless, if the AI findings do not accurately reflect reality, it is best to refine and iterate the prompt until obtaining the expected result. Saving Changes...
Saju DevassySenior Program / Project Manager| DELLEMCBangalore., Karnataka, India
There is nothing to Worry about the output, unless the inputs is Rubbish, You get what you give it (AI) ! IF your inputs are structured and well scripted (Prompts) your output will be good. The second word in AI that is the "Intelligence" comes in from the common sense that is fed is either by a "Human" or another "Agent" who was / is feeding data via an Human input.
Most if the AI outputs are always quoted at the end "This is an AI generated output, please excercise caution" when using it.
It again goes back to the common sense from the Human to determine to use it or No, Well AI makes life easier, but making sure the AI is making it easier for you remains with the individual using it. Saving Changes...
"But the fact that some geniuses were laughed at does not imply that all who are laughed at are geniuses. They laughed at Columbus, they laughed at Fulton, they laughed at the Wright brothers. But they also laughed at Bozo the Clown."
PMI Team Member
