Graphic design is ever evolving, and generative artificial intelligence (AI) is quickly becoming an important companion in the creative design process.
If you haven’t yet complemented your design workflow with AI, it is now easier than ever to start creating unique and vibrant designs to draw in your audience and elevate any story.
I wanted to give an outline of why to use generative AI to transform your design process and provide a practical example of how I used generative AI to complete a creative design brief.
What is generative AI?
Generative AI comprises artificial intelligence models that generate new content, such as text, images, music, or other media, based on their training data.
These types of AI models leverage machine learning algorithms, particularly deep learning techniques like generative adversarial networks (GANs), to analyze vast amounts of data and identify patterns. By learning the underlying patterns and structures in the training data, generative AI models can create content that is both realistic and novel.
Some of the more well-known examples of generative AI models include OpenAI’s Generative Pre-trained Transformer (GPT/ChatGPT) for text generation or Stability AI’s open source Stable Diffusion AI for image generation.
Why use custom generative AI graphics in your designs?
Generative AI has the potential to transform graphic design work in cool ways:
Instant content generation: Graphic designers can use AI models to quickly whip up visual elements, such as illustrations, patterns, and graphics, saving time on repetitive manual work.
Supercharged content quality: AI-generated content can be more layered than human-created content because AI models dig into into vast amounts of data and uncover patterns or layers that humans may not see.
Endless content variety: AI models can generate diverse eye-catching content of different types and styles allowing designers to cater to a broader range of tastes.
Personalized content creation: AI models can generate personalized content based on individual preferences, interests, or memories, delivering highly customized visual results.
How to start using generative AI for your graphic designs?
Consider the following steps to get started in including custom generative AI graphics into your design process:
Step 1: Choose an AI model that works best for you
Select a generative AI model with an interface that suits how you like to learn. Also, consider how much money you want to invest in the AI tool. Here are some popular AI models to consider (not affiliated, just examples) with a couple of descriptive comments:
Open AI GPT (ChatGPT-4) (text generation): Free tier available, web interface — Use this AI model to generate text-based content for your designs, such as prompts, headlines, captions, or descriptions.
Open AI DALL-E-2 (image generation): Free credits to start, web interface — This AI model can transform text descriptions into images, allowing you to generate custom visuals based on specific themes or styles.
Midjourney (image generation): Limited free trial to start, Discord interface — Use this model to generate rich images in various styles from intricate prompts. Harder to master but great image outputs.
Night Café (image generation): Free credits to start, web interface — This site offers a couple of AI model to transform text descriptions into images. It also has neural style transfer options to stylize your own images.
Jasper AI (text and image generation): Limited free trial, pricey, web interface. Great for marketers and content writers who write a lot and need hero images. Organizes content and has templates.
If you are already paying for a design program, keep an eye out for the release of new AI buttons and portals. Many companies are layering in new generative AI features into their software.
Don’t add a subscription when you don’t have to!
Step 2: Fine-tune the AI Model
After you have chosen an AI model, fine-tune the model to fit your design wants and needs and get more accurate content. Don’t give up too easily.
This process may involve playing with any settings provided with the AI interface or uploading additional data, such as examples of your previous designs or specific design elements you want to incorporate.
Step 3: Experiment with AI-generated Content
Start experimenting with the AI-generated content in your designs.
Mix it up: Use AI-generated images as backgrounds to make your hand drawn or original designs pop.
Add AI-generated words: Sprinkle some AI-generated text into your originals.
Multiple layers: Feel adventurous? Create an entire collage of AI designs using AI-generated elements. Use your design eye to stage a richly layered AI-rendered scene.
Remember that you may need to edit or modify the AI-generated content to better suit your design goals and maintain your unique identity. This leads us to step 4.
Step 4: Combine AI generated imagery with Other Design Programs
Once you’re committed to take the dive into the world of generative AI for graphic design, make any AI generated content more your own by editing the design with other programs like Procreate, Affinity Designer, Adobe Creative Cloud, or Canva — to name a few.
You can easily fuse AI-generated content into your style by blending AI generated content into your regular design method. Take advantage of your favorite design tool by spicing up your process with the power of AI elements.
A practical application — generative AI & the creative brief
Let’s test out the previous four steps. I created a dummy brief to provide a practical example of how a design process can be beefed up by AI.
🖼️ Project: Cake lover’s birthday thank you card
Objective:Create a vibrant and heartfelt thank you card for a cake lover, celebrating their upcoming birthday. The card should convey appreciation and gratitude while incorporating elements that reflect their love for cake. The goal is to delight the recipient and make them feel special.
Extra considerations: Use visual elements that evoke joy and celebration, use an appetizing yet happy color palette, include cake, use an expressive font, include an expressive message, make sure the layout is balanced between creativity and legibility.
Timeline: Due tomorrow (That long!!😄)
Deliverable: Digital or printable card
Step 1: I chose Midjourney (MJ) to prompt for ideas
My preferred generative AI program is MJ. I just like all the different ways MJ can be used to generate images. For now, a user needs a Discord account, but a web-based interface is coming. I usually do a lot of playing around with MJ versions and prompt qualifiers, but for this exercise, I kept the prompting simple.
Prompt: cake lover’s birthday thank you that captures a passion for cake
Step 2: Fine-tune MJ outputs
I started with my prompt above using the MJ “/imagine” command just to see what I would get as a baseline. Next, I started to change model versions using the “/settings” command. I also played around with the wording of my prompt. As I stated early, I did not go deep into prompt qualifiers (saving that for future projects).
Here is a collage of MJ AI images from my prompt(s):
Step 3: Mashing-up MJ AI content /Step 4: Combine AI generated imagery with other design programs
Next, I took inspiration from the MJ ideas and started reworking some of the images using Adobe Photoshop. I have a subscription to Adobe Creative Cloud, but there are many comparable software programs that cost less (i.e., Affinity Designer, Canva, etc.).
I combined two, reworked extraordinary cake images above with a free Adobe Illustrator postcard template and made some custom copy for my card (see elements below).
The Result — You take the cake, birthday/thank you card
🎂 The resulting card is a mashup of multiple elements that creates a birthday/thank you card for any of the cake lovers in your life.
Obviously, there is much more that could be done, but the whole process took about an hour. The imaginary client loved it!
What I wanted to show is how generative AI images can be used to mash-up with an existing workflow.
Using generative AI to create some of the card elements is not much different than customizing stock photos or vector templates.
What makes this design mine is that I created the blend of elements. The design fits with the brief and my inclination to make picture collages with words.
I hope my example helped show how incorporating generative AI into your design process can help with creating unique, dynamic elements.
By using AI-generated content along with editing by popular design programs, you can unlock new creative possibilities and streamline your design workflow.
Try embracing the power of generative AI to upgrade your design process, but remember to always keep your style along the way.
If you like this article, you might also like this one: Find your own style using Midjourney’s blend mode
Julia Fletcher is founder of JEFS Storytelling Arts, a graphic design studio, where she uses her unique research skills and artistic talents to create custom visual stories that help clients’ increase engagement and promote the education of their audience.