Welcome to AIxCreative Newsletter. Today is all about practical, quick workflows. Here is what’s on deck:
Workflow 1: Illustrator to Midjourney
Workflow 2: New Krea Flux Model & Ideogram for Packaging + R&D Design
Tips: Prompt of the Day
01 / WORKFLOW
Illustrator to Midjourney
We have a new workflow to share with you! This process is ideal for CMF experimentation on an existing form. The steps are as follows:
Create rough shapes: Start by creating an underlay in your preferred tool. In this example, I used Adobe Illustrator’s Image Trace tool to generate the underlay.
Export and input: Use the underlay as input for MidJourney’s Edit feature.
Retexture: Apply various style reference images and prompts to achieve different results.
Upscale: Finalize the design by upscaling in Magnific.
We like this approach because it allows the designer to focus on proportions, profiles, and design elements, blending human creativity with machine efficiency.
Here’s another example where Hector recently used this method to experiment with soccer cleats.
This flow results in a super wide range of possibilities while keeping time investment low. Try it with your own product in mind and let us know how your thoughts.
Clare
02 / WORKFLOW
Krea Flux Model & Ideogram for Packaging
We recently shared a packaging workflow on LinkedIn, showcasing how we moved from concept ideation to exploring various styling and branding options for a tequila brand. The results were solid, offering a broad range of CMF possibilities.
Here’s another example I put together for eyeshadow palettes:
Krea's Flux / Ideogram: Used for initial concept generation.
MidJourney External Image Edit: Applied for label variations.
MidJourney External Image Retexture: Used to explore additional concept directions.
While Hector achieved great results with his footwear example, this eyeshadow palette experiment didn’t go quite as smoothly. Although the outcome is still intriguing, we found that MidJourney’s Retexture feature struggled to maintain the design integrity of the original concept—especially with typography. (And yes, we’re aware the Japanese characters here are nonsensical!)
Learning by Doing
This is a great example of why we emphasize learning by doing. The nature of Ai today means that results can vary significantly depending on the product or category. Each project might require a different approach or set of tools, and the only way to truly understand this is through hands-on experimentation—what we like to call R&D Design efforts.
Around 70-80% of our work falls into the category of R&D Design efforts. These are “by any means necessary” projects, where we tackle real-world design problems, discover workarounds, and uncover ways to maximize the potential of these rapidly evolving tools.
In our experience, the main advantages of R&D Design efforts are:
Understanding the limits: You’ll learn to differentiate between running out of skill and running out of tool capability, which can greatly reduce frustration.
Uncovering possibilities: You’ll gain insight into what’s achievable with skilled Ao operators, helping you understand the true potential of these tools.
Building credibility: When R&D efforts are executed by a mix of domain experts and experienced Ai operators, workflows become more realistic, and results more credible—helping teams buy into the process. R&D efforts serve as proof-of-concept for Ai integration.
Despite the buzz around Ai, there are no shortcuts to becoming an expert. Headlines like “Become an Ai Expert in 5 Days!” oversimplify the reality.
Mastering these tools requires time, overcoming cognitive barriers, and understanding their long-term implications. However, when implemented thoughtfully, we’ve seen these efforts deliver value within organizations.
We’d be happy to discuss how our team of Ai explorers and experienced designers can help your organization understand Ai in a practical, impactful way.
Clare
03 / TIPS
Prompt of the Day: Style Reference Image
After quite a long break, Prompt of the Day is back!
Here is the Midjourney prompt for this modern picnic basket concept.
<style reference image> modern picnic basket, product design, contemporary materials, isolated white background --c 10 --ar 1:1 --s 10 --sw {50, 100, 250, 500, 750}
This is the style reference image I used:
The main takeaway here is that this super simple prompt gets out of the way. It’s general enough to provide a framework for the reference image to do all the talking.
I don’t have to differentiate between a lid or top and a body because my reference image is already showing a multipart object. As we discuss in our Midjourney for Design course, this is a prime example of talking with images.
Speaking of which, our Black Friday promo is still going on our eLearning modules. Head over to our website for the details. Those modules are a great way to get your feet wet in the Ai waters. Link here.
Until next time,
— Clare and the AIxC Team
Out of curiosity, why not use the upscaler native to MidJourney?