Looks like we’re not in Kansas anymore. AI and image generation.

dorthy and her little dog toto from the wizard of oz was the tv show meets artificial intelligence which is portrayed as a fantastical machine partly hidden behind a curtain, carpetpunk, light green and dark gold, high gloss --ar 50:21 --v 5

The rise of AI-generated images has taken the creative world by storm, with numerous AI tools and platforms generating breathtaking visuals that are quickly gaining popularity.

As a marketer in this wild new frontier needing quality imagery, this tool saves a lot of time and expense. And it’s mind-blowing what it can do!

As I’ve updated my website, new tools of the trade, like GPT4 and Midjourney, are rolling towards us like a boulder in an Indiana Jones movie; I thought I’d share some resources and prompts that are used on this site.

First, Midjourney has been my go-to tool for image creation. To use it, you’ll need to join its Discord community. Discord has a desktop app that’s simple to set up but a little foreign at first. Using it along with a paid plan provides the opportunity to create your own “bot” and receive direct messages with completed images. This saves a ton of time, rather than using the free plan and having to scramble in public rooms to find where your image is rendered in the scrolling thread. To learn more about pricing and explore Midjourney visit this link.

So, let’s take a look at some of the prompts and the results.

For the home page, I felt it was important to use at least one hero image created by AI.

As it’s valuable for every startup to have your team rowing in the same direction. That was a great one to explore.

Here was the prompt: /imagine: rowing in the same direction — aspect 2:1

Pretty simple prompt and interesting results generated by version 4 of MidJourney. The images are numbered top to bottom, left to right. Meaning 1 is upper left, 2 is upper right. The rectangles below the images with “U” are to upscale that image, and the “V” is your asking for another revision based on that image. The blue button with arrows gives it another pass without referencing the last images.

Clicking “U3” generates an upscaled version creating a 1728px wide image. Version 4 has a beta upscale, which can be applied to an upscaled image taking the resolution up to 2048px. Version 5, the latest to date, will only upscale to 1024px. However, the image’s artistic quality is so much better than V4. More on the different versions can be found here.

So in this instance, I liked image V3 the best; here are the four variations that were generated. Clicking on the image opens it in the browser, making it easier to view.

I selected V3 as the final one to upscale. As you can see, it did some funky things to the oars, and with hands, even V5 of Midjourney is still pretty bad, extra fingers, not enough fingers. Just weird. After downloading it in Photoshop, I converted it to grayscale to fit the aesthetics of my brand.


The art of prompting.

Check out Nick St. Pierre on Twitter to see just how impressive AI images can be. Here’s one he prompted by combining a single prompt in four languages. The entire prompt can be viewed here.

Nick St. Pierre

As AI-generated images have taken the creative world by storm, questions arise regarding their impact on the stock image industry and the copyright challenges it presents.

Concerns for the Stock Image industry.

Infinite Supply of Unique Images: AI image generators can create an endless supply of unique, high-quality images on demand tailored to specific needs.

Reduced Production Costs: Creating AI-generated images often involves lower production costs than traditional photography.

Customization and Personalization: AI-generated images allow for greater customization and personalization, enabling users to create visuals that closely align with their brand identity or artistic vision.

Copyright-related challenges.

Ownership and Attribution: Who’s the creator/owner of the image? The AI algorithm, the user who generates the image, or the AI itself?

Infringement: When training the models, the AI soaks up everything it’s given. In Midjourney’s case, that was copyright-protected images as well as the AI algorithms draw inspiration from an image. For example, I can put an original image for direction in the prompt. This raises questions about liability and how AI-generated images can be considered original works.

Fair Use and Licensing: The widespread use of AI-generated images may lead to new interpretations of fair use and licensing. For instance, the licensing terms for using AI-generated images may need to be reconsidered. New guidelines need to be established to ensure that creators and users are fairly compensated for their work.

The rapid advancements in AI-generated imagery are incredible and terrifying at the same time. As things are moving so fast, it isn’t easy to feel that there’s a sense of certainty. Still, businesses, content creators, and legal experts need to stay informed and engage in discussions about the ethical and legal implications of this emerging technology. By doing so, we can work towards finding a balance between encouraging innovation and protecting the rights of creators in this exciting new frontier.

Other AI generative tools to explore:

DALL-E: The Picasso of AI Art

Adobe Firefly: Time will Tell

Artbreeder: Creative Collaberation

Generative Art Playground: Tech Playground

RunwayML: The Swiss Army Knife of AI

DeepArt.io: Unleashing Your Inner Van Gogh

NVIDIA Picasso: Coming Soon

These tools are revolutionizing the creative process, and I’m excited and terrified at the same time to see what the future holds. We’re definitely entering the land of OZ.

Midjourney AI-generated images on this site:

Follow on LinkedIn


Get in Touch

Or, if you’d prefer to talk rather than text!