The Art of Prompting

In the modern AI-enhanced workplace, it’s likely that you’ll find yourself leveraging generative AI daily. Generative AI involves communicating with AI through prompts. The art of prompting stands at the intersection of human intention and artificial intelligence capability. While AI language models possess remarkable capabilities, they are fundamentally dependent on how well we communicate our needs and expectations. A well-crafted prompt can mean the difference between receiving generic, unfocused output and getting precise, actionable results that directly address your objectives.

Prompt engineering

Prompt engineering is the practice of crafting and refining prompts to effectively communicate with and get optimal results from AI language models. It involves carefully structuring your questions, instructions, and context to guide the AI toward producing the specific type of response you want. Think of prompt engineering like learning to be an expert interviewer or teacher - you need to know how to phrase things clearly, provide relevant context, set appropriate constraints, and guide the conversation toward your desired outcome. This could involve techniques like:

• Breaking complex tasks into smaller steps
• Providing relevant examples or templates
• Setting clear parameters and constraints
• Specifying the desired format or style
• Including necessary context and background information

Good prompt engineering requires understanding both the capabilities and limitations of AI models. It's about finding the right balance between being too vague (which can lead to unfocused responses) and too restrictive (which might constrain useful output). The field continues to evolve as AI models become more sophisticated, but the core principle remains: better prompts lead to better results.

For professionals navigating the AI-enhanced workplace, prompt engineering isn't just a technical skill—it's a core competency that enables you to effectively delegate tasks, solve problems, and amplify your productivity. Whether you're using AI for content creation, data analysis, coding assistance, or strategic planning, understanding the principles of effective prompting will help you consistently achieve better outcomes while avoiding common pitfalls that lead to suboptimal results.

Explore prompting with an AI assistant

While there's various generative AI tools available these days, we'll start with a simple AI-assistant. An AI-assistant is a conversational interface powered by large language models that can understand natural language inputs and generate human-like responses. Examples include Claude (from Anthropic), ChatGPT (from OpenAI), Gemini (from Google), and Copilot (from Microsoft). For this exercise, you can choose whichever assistant you prefer from the list – do note that the ones provided have a free version available.

Let's begin by asking the AI to help create a professional email explaining a project delay to a client. This is a common workplace scenario where AI can help craft a well-structured message.

Here's an example prompt you could use:

"Write a professional email to a client explaining that our project delivery will be delayed by two weeks due to unexpected technical challenges. The client's name is Eric, and the project is a website redesign originally due December 15th. Maintain an apologetic but confident tone, and include specific next steps."

Results can vary significantly based on the AI assistant you use and how you phrase your prompt. Here's the response I received:

Subject: Project Update: Website Redesign Timeline Adjustment

Dear Eric,

I hope this email finds you well. I'm writing to inform you about an important update regarding your website redesign project.

Due to some unexpected technical challenges we've encountered during the development phase, we need to adjust our delivery timeline by two weeks from the original December 15th deadline. Our team has identified the issues and has already implemented solutions, but ensuring everything meets our quality standards will require this additional time.

To keep you fully informed, here are our next steps:

1. Complete resolution of technical issues by December 20th

2. Conduct thorough testing by December 25th

3. Deliver the final product by December 29th

We will provide bi-weekly progress updates and are happy to schedule a call to discuss any concerns you may have.

We truly appreciate your understanding and remain committed to delivering a website that exceeds your expectations.

Best regards,

[Your name]

You can try additional prompts as well and compare across assistants – which is a key skill you'll learn to foster as you explore AI tools. For example, you might try:

• Asking for a more formal or casual tone
• Requesting bullet points instead of paragraphs
• Specifying different project circumstances
• Adding more context about the client relationship

This hands-on exploration will help you understand how different prompts and AI assistants handle the same task, allowing you to develop better prompting skills for your specific needs. 

Building your prompt library

A prompt library is your personal collection of tested, refined AI instructions that consistently produce the results you need. Think of it as your AI playbook - a growing resource that makes your interactions with AI tools more efficient and reliable over time.

The foundation of any good prompt library rests on these key principles:

Clarity of Purpose

• Start with a clear objective statement
• Specify desired output format and style
• Include relevant context and constraints
• Define success criteria

Structured Format

• Use consistent formatting patterns
• Break complex requests into discrete steps
• Include explicit instructions for handling edge cases
• Maintain clear separation between context and instructions

Contextual Richness

• Provide relevant background information
• Include necessary domain knowledge
• Specify audience and tone requirements
• Define technical or industry-specific terms 

Output Control

• Specify desired length and format
• Include quality criteria
• Define error handling preferences
• State any required citations or attributions

Remember that your prompt library should be a living document that evolves with your needs and experience. You’re encouraged to test the prompts provided across the various AI assistants shared in this chapter. Who knows, maybe you’ll find your favorite AI assistant in the process!

There's more than one way to prompt

What I've shared with you thus far is just one of many ways to approach prompt engineering. As you explore other resources that focus on prompting, you'll likely be introduced to other strategies and techniques to follow. My best suggestion is to experiment and choose an approach that works best for you!

More About the Series

This article is part of the Resolution Proof Your Career: Thriving in the AI Workplace series - a 6-week career transformation series designed for professionals who want to turn their 2025 AI career resolutions into reality - starting now, not January. It’s not too late to join! Subscribe to the Millennial Workweek newsletter to get the next exercise in the series and visit the Millennial Workweek website to get caught up!