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?
Marcos CamposSenior Project Management| Andes DigitalSantiago Centro, RM, Chile
In my experience the key relies on being specific as you can, and give all the documents and examples as you have. That would short the iteration loop. Saving Changes...
INDUMATHI KANNAYIRAMPROJECT MANAGER| DELTASTAR POWER PROJECTS SERVICES LLCAbudhabi, U.A.E, United Arab Emirates
font _msthash="385" _mstmutation="1" _msttexthash="404144"1: Writing Assistance/font
font _msthash="387" _mstmutation="1" _msttexthash="511199"2.Technical Explanation/font
font _msthash="389" _mstmutation="1" _msttexthash="653133"3.Customer Support Response/font
font _msthash="392" _mstmutation="1" _msttexthash="337402"4.Creative Writing
font _msthash="393" _msthidden="1" _mstmutation="1" _msttexthash="9607"5./font/fontfont _msthash="394" _mstmutation="1" _msttexthash="261183"Marketing Copy. I/fontfont _msthash="396" _mstmutation="1" _msttexthash="32153186"n each example, the improvements involved clarifying the prompt's intent, specifying target audiences, and highlighting key elements to focus on. These changes led to outputs that were more accurate, relevant, and aligned with the original goals./font
Crafting Clear Prompts and Define your goals: Be clear and specific about what you want to achieve with the AI system. What information are you looking for? What task do you need it to perform?
Use concise language: Avoid ambiguity and overly complex sentences.
Provide context: Give the AI system enough background information to understand your request.
Be critical: Don't assume everything the AI system generates is accurate. Fact-check information and verify its credibility using reliable sources.
Consider alternative outputs: If the initial results aren't what you expected, try reformulating your prompt or exploring different functionalities offered by the AI system.
Provide feedback: If possible, give feedback to the AI system about the quality and relevance of its outputs. This can help improve the system's performance over time.
Important to receive the results desired:
We need to be precise and clear.
Be clear with jargon or specialized terminology.
Provide a clear context to all requests.
Important that outcomes are a consequence of what we input to the LLM
Not being scared of experimenting to obtain better results. Saving Changes...
Anonymous
A good way to validate information from AI is to check for assumptions underlying the response. Secondary checks on the credibility of resources used by the AI should help to guage confidence in the reponses. Saving Changes...
In my experience with AI, I start by defining a clear persona prompt to establish the context and expected outputs. I then iterate through multiple versions, adding specific context and refining the prompts based on initial outputs. To ensure accuracy and relevance, I implement strong testing protocols, comparing outputs against established criteria and original goals. Continuous interaction and feedback loops help in achieving the desired outcome, ensuring that the results are accurate, relevant, and aligned with my objectives. Saving Changes...
Be precise and clear while giving instructions.
We can use methods like RTF CREATE etc, as per our use case to generate appropriate output.
Saving Changes...
Bledar BeqiriProject Manager| Swedish Institute for StandardsStockholm, Sweden
It has helped me identifying the GenAI-system as a PM colleague. How do I interact with my colleague(s)? I ask clear and concise questions, give feedback, refine the information given and ask again - until the original goals are satisfied.
In short, be as diligent and aware of the AI as you are with a human counterpart! Saving Changes...
PMI Team Member
