Exploring AI Image Generation: Learning to Prompt Creatively - Bonus Activity

    This bonus activity gave me the opportunity to better explore a different side of AI, which is image generation. This was my first time experimenting with tools that turn text into visuals. I explored both Microsoft Copilot (which uses DALL·E) and Google Gemini to see how each responds to image prompts. The general thought is that generating images using AI is simple. Just describe what you want and AI gives you the result. The truth is, image prompting requires a different kind of thinking compared to text prompting. 

    My first attempts were very basic, such as: “Create an image of a workplace training session.” The results were acceptable but generic and not very engaging. The images lacked detail, context, and creativity. They looked like the generic images you can find on the internet and not really related to the context I was concerned with. Then, I refined my prompts and started adding:

- Setting (specific organization, office, classroom, virtual environment)

- Style (realistic, illustration, modern design)

- Details (program name, diverse employees, interaction, technology use)

- Purpose (training, engagement, collaboration)

    The results improved significantly. The images became more aligned with what I had in mind and more useful for real applications like presentations or communication plan.

Tips and Insights from My Experience

Through this process, I learned several important lessons about prompting for images:

1. Be descriptive, not vague.

2. The more details you include, the better the output you get.

3. Specify style and tone. This makes a big difference in how the image looks.

4. Think visually and imagine the scene clearly and describe it as if you are guiding a designer.

5. Refine your prompt. The first prompt will rarely give the best result.

6. Different tools produce different result. I noticed that Copilot tends to create more polished and artistic visuals, while Gemini can be more experimental depending on the prompt.

My Prompt and Generated Images

Tool Used: Microsoft Copilot (DALL·E)

Prompt:

Create a realistic and professional image of a corporate training session where a diverse group of employees are actively participating in an interactive workshop. The setting should be a modern office with digital screens, collaborative discussion, and a trainer facilitating the session. The mood should be engaging, positive, and focused on learning and development.

Generated Image:

Image generated by Microsoft Copilot (DALL·E)

Tool used: Google Gemini (Nano Banana)

                                                 Image generated by Google Gemini (Nano Banana)

    Prompting for images is a technical and creative skill. It requires clarity, attention to detail, and the ability to visualize ideas before describing them. For HRD professionals, this skill can be useful for:

- Creating training materials

- Designing presentations

- Enhancing communication campaigns.

Comments

Popular posts from this blog

From Slides to Simulations: How Technology Is Reshaping Workplace Learning

Building a Personal Learning Network as an HR Professional

What Makes a Blog Engaging?