Kicking off a conversation with a computer is akin to initiating a dialogue with a human. Just like setting an event on someone’s calendar, it’s best to provide context and a clear objective to avoid confusion with AI.
When starting the conversation, be sure to provide a clear structure. Let the model know precisely what the conversation’s objective, or desired outcome, is. Set an agenda, just like you would with a meeting. Provide clarity around what needs to be addressed.Â
Similarly, provide any relevant contextual information to help the AI understand its role in the conversation and yours. And provide lots of background information. Remember, more is more. Be sure to keep your framework consistent throughout the discussion.Â
Specifically, input any data or details the AI will need to process and analyze. For example, if setting quarterly objectives, you might provide financial statements or employee morale surveys.
Engage in an iterative feedback loop with the model (also known as recursion) and employ retrieval-augmented generation for added context as required.
For effective interactions with AI, give direct feedback. Don’t worry – AI doesn’t have feelings! Ask the model to reflect on its answers, and instruct the model to ask for more information when needed.
Goal: Clearly define what you aim to achieve.
Context: Provide all relevant background information.
Persona: Describe the ideal expert for the task. (Efficacy of this is still debated.)
Instructions: Provide clear, step-by-step guidance.
Gold Standard Examples: Offer high-quality examples of desired outputs.
Hyper-growth Mindset: Combine an ambitious goal with a humble approach.
Prompts to Stimulate Reasoning: Engage the model in critical thinking.
To engage effectively with AI, provide a clear goal ample context, and approach the interaction with a growth mindset. As AI continues to evolve, many advanced features will become standard, streamlining the conversation process. Always be open to asking questions and iterating for better results.