Director, Learning Design & Development| PMIAsheville, NC, United States
With Generative AI, iteratively refining and optimizing prompts can lead to better AI-generated results. This may involve adjusting the specificity or clarity of the prompt to increase relevance and accuracy of results.
What examples do you have of how improving a prompt drastically changed the output quality? What specific changes did you make that led to the improvement?
I like to phase my prompts in a series of steps and often map / plan my prompts before engaging with the AI tool. This saves time and less rework in revising one prompt over and over again.
I like to phase my prompts in a series of steps and often map / plan my prompts before engaging with the AI tool. This saves time and less rework in revising one prompt over and over again.
I do the same. Saving Changes...
Derek JonesCompliance Manager| BankingPa, United States
Jun 21, 2024 9:36 AM
Replying to Eduard Hernandez
...
Increasing specifity and more context leads to more accurate and refined results/output. I am not familiar with the diverse frameworks provided by Sergio Luis Conte; speaking to GenAI engine as I would speak to another human (thus, providing context and sufficient level of detail) provides great outputs.
That is a great point to converse with the AI as you would a person. With each piece of feedback and information, you can synthesize your understanding and where you need to go. Using a model such as Action Learning, etc. would be ideal for this process. Saving Changes...
Mahboobullah SeddiqiSenior Adviser| Ministry of EconomyKabul, Kabul, Afghanistan
As a project manager, I’ve seen how tweaking a GenAI prompt can totally change the game. A vague prompt gets you generic, meh answers, but a well-structured one, where you’re clear about the context, goals, and details; gives you spot-on, actionable insights. It’s super helpful for things like risk planning, keeping stakeholders in the loop, and making smarter project decisions. Saving Changes...
The more specific you are with the prompt the more accurate the data will be. I recently used effective prompt analysis to get AI help with a research topic within the financial sector. When i made the initial request, i received general topics which were vague however after carefully tailoring the request to a specific aspect i got the perfect topic for the research Saving Changes...
In a nutshell, prompt engineering has saved time and improved the quality of my AI generated results. When I first started using LLMs I knoew nothing about prompt engineering. Thankfully I stumbled on the topic while watching a YouTube video.
Now, I never use GenAI without first thinking of the prompt...second nature, even for simple things. Saving Changes...
Babatunde FakunleExecutive Director| Centre For Sustainable Access to Health in AfricaStoney Creek, Ontario, Canada
I use this six-part prompt formula that I saw on YouTube and it has never failed me: task, context, exemplars (example), persona, format, and tone in that order of importance, starting with the task. It is similar to the CREATE formula mentioned in this course. Saving Changes...
font style="vertical-align: inherit;"font style="vertical-align: inherit;""Como membro do PMI enquanto uso IA para pesquisas em gestão de projetos e outras atividades científicas, descobri que o refinamento de prompts impacta diretamente a qualidade das respostas geradas. No início, meus prompts eram genéricos e um pouco estruturados, o que resultava em respostas superficiais. À medida que aprimorei uma técnica de refinamento, as respostas passaram a ser mais precisas, aplicáveis e contextualizadas./font/font/font/font Saving Changes...
Moustapha NdiayeCoordination officer Partnership and Development Finance| UNITED NATIONSDakar, Senegal, Senegal
Outputs are more precise Thank you Saving Changes...
As a communicator, we always check that the message is being understood as intended (decoding). If the message is very complex, we look for ways to break it down to make it understandable, and we verify that each of its parts is understood. We should treat AI in the same way, relying on the rules and formulas that have been transmitted to us. I've had cases where, because the message was very ambiguous or imprecise, it was difficult to get a response that fit the logic we expected. Saving Changes...
PMI Team Member
