Midjourney and AI

With all the AI hysteria I decided it was about time I gave it a try myself. My initial experiments with ChatGPT left me extremely impressed with the new tool. Definitely saw some opportunities to have it write copy for me. Planning to leverage that more in the future for sure!

One of my first prompts was able to produce this! Generated using: “Hyperrealistic rhino silhouette on the african plain at sunset”

In addition to that I also payed the $10/month to Midjourney to try generating some images with the same power I witnessed behind ChatGPT. It was a fun exercise to see how detailed I could go, how it would adapt to different style prompts, and just generally get a feel for how image generation AI works. This also let me see some of its current potential limitations or areas I may need to work more on prompting to truly get the results I wanted.

I don’t really want to dive into the morality/legality of these AI images at this point, but I suppose I could still talk around potential use cases for a tool like this. This tool has worked great for me when I only have a loose idea of what I want to see. I also think it works amazing as a jumping off point or serving as inspiration for a new project.

Generated using: “Steampunk mechanical dog, cogs, detailed, hyper realistic, cinematic”

Often if I have a clear image in my head, that is not what I will see in my image generations. This is especially noticeable when trying to craft a chart or new UI. Other common issues I’ve seen are around any sort of generation of text. Without changes I think the need for designers will continue to remain and even then in order to get to a hyper-specific solution I think it still makes sense to create something by hand.

However, for things like blog banner images or content images I could see a tool like this be extremely helpful for saving time. Overall, I think this new tool is incredibly exciting. I don’t think designers need to be intimidated, but more learn how to utilize this assuming it becomes a part of the design industry. From what I’ve seen on other design forums, this seems like a pretty common sentiment.

Personally, I do feel a lack of “ownership” on images generated by AI. They don’t feel like they’re really my creation. (Maybe somehow related to IKEA effect haha) For now I’ve been using them to generate avatar images for my profile picture on sites like Discord.

Generated using: “Avocado and watermelon as one fruit”

Leave a Reply

Your email address will not be published. Required fields are marked *