I have always been interested in art, and I love the way that it allows you to express yourself. I have always been fascinated by the way that artists can create something so beautiful out of nothing. I am also fascinated by the way that art can be created using Artificial Intelligence, and I believe that the AI art generation is the future of art.
There are a number of AI art generation tools available, each of which offers its own unique capabilities. Some of the more popular AI art generation tools include Dall-E 2, MidJourney, and Stable Diffusion. These tools allow users to create unique images from textual descriptions.
In this article, I’m going to show you my first AI-generated images using Midjourney and explain how you can get started with creating your own.
My AI-Generated Art Pieces
Here’s some example of images I created in a few seconds. Check out this page to see what else I’ve created using the Midjourney AI program:
Video Tutorial on Getting Started with MidJourney
If you prefer to watch a video to learn about how you can get started with AI Art generation using Midjourney, then here’s the video for you:
Accessing MidJourney
In order to access Midjourne, first, you need to install Discord, which is a free instant messaging social platform. Create an account for yourself and login into the Discord application.
Then go to Midjourney.com website, and click on ‘Join the Beta’. Currently, Midjourney is in Beta.
It would redirect to the Discord application and once authorising is done it will open the Midjourney channel in Discord.
Imagine Something and Prompt the AI to Generate it!
In order for Midjourney to start creating AI Art, you need to navigate to the ‘newbies’ chatrooms in the left sidebar of the Discord channel. Once you are in the channel you’ll see real-time activities as others are trying various commands for AI to generate images for.
To ask the AI to draw something you’d need to type the ‘/imagine’ command and type a prompt to explain whatever you like the AI to create (as shown below).
Once you enter the command, you’ll see the image consisting of 4 variations being generated as the below image:
For example, I asked the AI to generate an image of a “nebula in the shape of a waterfall”. Below is the first results I got. The AI program generates a 2×2 grid image with 4 variations of what you prompted it.
You have the option to re-generate the images if you didn’t like any of them. Simple click on the last button with the icon with the circular arrows as shown below. If you like any of the 4 generated images, you can upscale the image to create a higher quality version of that variation by pressing the U1-U4 buttons. The top-left image is number 1, the top-right image is number 2, the bottom-left is number 3, and the last one is 4. If you like a particular image but like to see more variations of that then you can click on the V1-V4 buttons.
In this case, I created some variations of the first image and It came up with the following picture.
Then I upscaled image 2 of the previous variation, and the AI came up with the following detailed image.
With the upscaled image, you have the following options to modify it further. You can create more variations from it by pressing ‘Make Variations’, or upscale it further to get a higher resolution image by pressing the ‘Upscale to Max’. You can also reduce the details of the image by pressing the ‘Light Upscale Redo’ afterwards.
How to Find Your Generated Images Quicker
As you try different commands, you might notice that the screen jumps up and down because other people are trying to generate images at the same time in the same chatroom and it could get a bit crowded. Once you purchase a plan then you get a dedicated chatroom that you can use exclusively without interruptions.
One tip for finding your generated image quickly is to go to the Inbox folder (the icon is located at the top-right side of the Discord channel) as highlighted in the image below, and then click on the Mentions section. This will show all of your recent works and by hovering over them and clicking on the ‘Jump’ button you can jump to your generated image quickly.
You can view all of your generated images in the profile dashboard as shown below:
You can save and download the images. You can also check and copy the prompt that you used previously in order to see how you generated a particular image and regenerate it.
Once you have paid plan you will get access to more features.
Midjourney Subscription Plans
Midjourney allows you to use 25 min of the fast GPU time for free. This is almost equivalent to running 25 jobs. A job is referring to either using the ‘/imagine’ prompt command to generate new images or using the buttons to create a variation (V1 – V4) or upscale (U1 – U4) of the generated image or the other buttons as discussed above. Once you run out of the free jobs your Free Trial period will finish and you would need to purchase a subscription. You would get a message: Free Trial Complete: Your free trial has come to an end. Please consider subscribing to one of our plans. To view more details, use the /subscribe
command or visit our website!
There are two main plans for individual use:
- The Basic plan, which is $10/month, gives you 200 min/month of fast GPU time and you get access to the personal Bot chat. This means that you get your own private chat with the bot and don’t have to use the common ‘newbies’ channels that are crowded.
- The Standard plan, which is $20/month, gives you 15hrs/month of fast GPU time. You could also activate the Relax GPU time where you get unlimited usage but the jobs get done at a slower rate. It’s good if you want to generate some jobs to run when you are away from the desk.
The below image shows more details about the plans.
To find out how many free tries you have got left, you can use the ‘/info’ command. When you are ready to subscribe to a plan, type ‘/subscribe’ in the common chat or the Midjourney bot. This will give you a link where it would take you to the page to get your subscription plan.
Midjourney’s Community Showcase Portal
Midjourney has a portal where they showcase the AI-generated images created by different people. You can view the recent generated images or the top-rated ones. You can bookmark the ones you like. You can also view the prompt description used for each photo. This would give you some insights into what others used to create a particular image.
Conclusion
I am absolutely passionate about the Midjourney AI art generation tool. I love the way that it allows you to create unique pieces of art, and I find the process to be very relaxing. I hope to continue to explore the world of AI-generated digital art. I’m amazed by how creative AI can be and I would not have imagined what AI could do a few years ago. Creativity and art were assumed only for human beings but now machines can be creative too! It reminds me of the famous statement of Edsger Dijkstra:
“The question of whether a computer can think is no more interesting than the question of whether a submarine can swim.”
Let me know what your thoughts are about this amazing AI tool and what it can do!
Pingback: My First Attempt to Create AI-generated Art Using Midjourney
Pingback: 5 Ways to Earn Money from AI-Generated Artwork (DALL·E 2, Midjourney & Stable Diffusion) - Brainy Loop
Pingback: 5 Ways to Earn Money from AI-Generated Artwork (Midjourney) - AI Toolz: AI Tools & Tech