Biblical Scenes Depicted by AI Pt. 2

Six months ago (November 2022) I published a post with AI artwork generated by Midjourney depicting scenes from the bible. Those images were generated with the 3rd version of Midjourney and while those were impressive, I went back and generated similar images with the latest iteration, version 5. The fidelity of these images blows away what was generated just half a year ago.

The rate of advancement of this technology is stunning. These are just a few of the about 200 images I’ve generated with this 5th version over the last month as I’ve been experimenting with the tool. There is a vast amount of variation within generated images, which means that each image generated is the result of a “dice roll” effect that produces a one-off effect. Generate more images with the same prompts and you’ll get different results. Some of these are lifelike, and some look like paintings.

Click on any image to display a larger version.

Crossing the Red Sea
Crossing the Red Sea
Jesus stills a storm
Jesus stills a storm
May 4, 2023|

I Generated Hundreds of Paintings with AI

The last three weeks I’ve been generating a lot of art with MidJourney after reading an article earlier this month about how someone in Colorado won an art competition by submitting a generated piece. When you see the caliber of art which AI algorithms like MidJourney are generating, it’s not hard to see why it won the competition: the results are strikingly good. So good that I wonder what AI is going to be doing this well next month, next year, or the next decade. We were told that creative disciplines like art and music were the last stations which artificial intelligence would take away from human creators. That is seeming like either a naive or short-sighted take at this point.

As of the last week of September 2022, I’ve generated a total of 233 images, which have consumed 3.6 hours of computational server time. The cost to me has been a paltry $10/month subscription fee. The way the algorithm works is you input a prompt and the machine outputs four images, which you can request iterations or upscaled versions of. As you interact with MidJourney more you realize that it becomes like a game, or a dice roll, where you are asked to input an imaginative prompt which the machine can meaningfully reflect upon. The results are very unpredictable; the output can range from underwhelming to incredible.

Here are some of the images I’ve generated so far.

Futuristic Chinese city
Sichuan Earthquake
Ninja Scroll in the Mist
September 22, 2022|
Go to Top