It took me a while to get there. Not sure why I waited this long. But Midjourney is amazing!
Midjourney? Midjourney is an AI-based image-generation tool that, to me, harkens back to my days of discovering Photoshop 5. I never looked at creativity the same way after Photoshop. And tools like Midjourney are another pivotal point where we question and explore the true essence of creativity in art.
The set up process can be a chore. But once you are up and running, it is easy to use.
My use of Midjourney: I am, and will be using MidJourney to maintain a (near-) daily journal where I create/prompt images based on ideas and musings that strike me as interesting.
My first two entries are already posted on Threads:
Disclaimer and Backstory: I am no artist — I cannot hold a crayon to save my own life. And this where Photoshop was so very empowering. At the same time, I do not pretend to be an artist who can draw and create from a blank sheet of paper with a pencil, ink and some paint — I am not; and I will never be.
However, I am a huge fan of the arts. I can spend days on end thinking about the finer details of a creation — be it prose, poetry, paintings, or pottery.
And this is where Photoshop was so very liberating — I could for the first time, express myself in ways that true artists were freely able to. I could never compete with them (still cannot) in creating works of art. But I could put together a bunch of layers in Photoshop, and say, “this is what I am thinking of.”
This enabled me to work with some of them over the years. And in the process gain a deeper appreciation into the process of design, art, and creation in general. This informs my work as an engineer and as product developer. For instance, I am no longer satisfied with a functional piece of code — my lesson with artists and art taught me that my code and its architecture needs to be elegant.
Why Midjourney and GenAImagery? I am fascinated by Midjourney and GenAI-based image creation for a bunch of different reasons:
- First, these AI models for image creation are smaller than LLMs (like ChatGPT). This allowed me to tinker a little on Stable Diffusion (SD) — another image generation tool. While SD is really cool, Midjourney makes generated images look really good, and detailed (I will blog more on this aspect on a latter date, when in compare MJ and SD).
- Second, I am fascinated by the Prompts to these Image Gen models. If you put qualifiers like “wide-angle”, or “digital art”, or “cartoonish” in your prompt to these image models the results tend to vary – a lot! To think that we can just create variations of an idea — pictorially — by just adding artistic qualifiers like that is pretty amazing.
- Also, I have started to view this as a substitute to photography. Photography is a way for me to explore a subject with images. When time permits, I love hitting the streets and clicking photographs. But as a new parent, I barely have the time/energy. But to be able to explore a subject or concept in my head, by applying lessons from photography, without leaving the side of my baby’s crib … is a huge selling point.
The cover image for this post was the first ever prompt I sent to Midjourney: “a lone human looking from a distance at a valley of mystery and magic, fantasy, colorful, 4K, digital art, hyper realistic.”
What a time to be alive.
— vijay, enjoying a smoke-free sunset
Winter Rant 
Leave a comment