So after my last post about how I Got Got by AI Art, I went ahead and started exploring this tool called Midjourney AI. Long story short – coz it’s a bit complicated and I wanna get into the fun stuff – you join a private Discord chat with the “Midjourney Bot”, you type in “/imagine”, and then add your prompt. With that, voila, you’re making awesome, sometimes trippy, sometimes fun, sometimes annoyingly abstract art.

There’s a trial that’ll last you about 10 minutes if you really get into it and then it’s $10/month to continue using the program. I’ll get into this more towards the end.

Let’s get into the artwork…. There is some cool stuff in here, but I saved the best for last, so just wait for it!

First, I wanted to test how close the A.I. would match my immediate imagination, so I started with a prompt for an image idea I’ve had in my head for over a decade. You can see my exact prompts underneath all of the following images.

So if you look at the prompt, you’ll see that it missed the mark a bit, but this is pretty damn close if you think about it. Immediately I was hooked.

Now here’s a set of images of a character I’m working on for a book…

Then I did this as I learned a bit more…

Much later, I watched some videos on how to create repeatable characters. Here are my favorite images that I got…

Now this was very exciting. This is really how I pictured this character. I was able to keep creating these prompts and refine the look I want to then create a consistent look to the character. Now it’s not as simple as this may seem, but overall, the more you learn and the more the A.I. learns, the better your results get.

Okay, so those are the basics… let’s get into the wild stuff!

From here on, I just wanted to see what this thing could do so check out the prompts I did below.

I tried to combine the last two, but it didn’t workout that well so I tried to combine them in a new prompt.

There I misspelled the Jiminy for Jiminy cricket, but I think it got the idea.

An interesting feature is Midjourney A.I. is that you can take an existing image, input the image link, and then add prompts to that. So what I did is I took one of the above images and added “art deco, 32k”. Here’s what I ended up with after some trial and error.

Okay so I’ve got a little crush on Christina Hendricks.

I went to church and there was some really beautiful imagery from Exodus…. so I put it into the prompt and got some really cool results.

AND THEN THERE’S THIS!!!!

I took a photo from the internet of The Loftus Party’s editor-in-chief/fearless leader, Michael Loftus, and let it rip!!!

Now, it doesn’t look much like Michael Loftus, but I bet if I trained Midjourney for long enough, I could get a spitting image. Now here’s where the reality sets in…

I ran out of “Fast Hours”. Fast Hours is the time allotted to you for the A.I. will process your prompt request. At the base subscription of $10/month, I received 200 minutes of Fast Time per month. I ran out of that in less than a week. If you upgrade, to a $30/month subscription, you get 20 hours of Fast Time per month (I believe) and you have the option of using Relaxed Mode, which lets the A.I. generate prompts on it’s own time. With Relaxed Mode, you can make an unlimited number of generations, it just takes a bit longer. You cannot use Relaxed Mode in the base subscription which is a real bummer.

So… what do you think of the images “I made” using Midjourney A.I.? Worth another 20 bones per month to keep playing around?

I see some seriously epic use-cases for this tool, but it’s also really addicting and, damn, I’ve got a lot of subscriptions.

What do you think? Keep the A.I. Art Factory alive and continue to fund what might be our extinction? Or should I just move on and consider this a fun $10 experiment? DO TELL! Seriously, I need to know what you think!

Enjoy!