Here are a couple that I did recently, based on my own mnemonics:
笛 (ふえ) - flute, sounds like “full egg”
DALL·E 2 - “A man is balancing a flute on his nose. Balanced on top of the flute is an egg. In a bamboo field surrounded by bamboo.”
過去 (かこ) - past (mnemonic: in the past, my car was an adorable car 子)
DALL·E 2 - “old sepia tone cartoon of a car with big innocent eyes and a small body”
This king has a drop of water floating above his head. That’s because he’s not just a king, he’s the master of water. Master of elements. Master of ALL THINGS!
This is my favorit one:
Here’s another one I just did with my own mnemonic, to try to better fit the image:
載 (さい) - publish (mnemonic: a cyborg (さい) with a giant cross on his car is driving while drunk, and just discovered someone published a story about his reckless behavior)
DALL·E 2 - “A newspaper article about a man driving while drunk. There is a giant cross attached to the hood of his car. A cyborg is reading the newspaper article.”
One more:
応 (おう) - respond (mnemonic: there is a king (おう) under the canopy of hearts, and if you write him a love letter, he will always respond)
DALL·E 2 - “A king with a crown is sitting in a small tent. He is writing a love letter covered in red hearts.”
Please anyone give me a list of about 10 requests for tomorrow morning 
I’ve not joined the midjourney beta, but I’d be extremely surprised if all of these queries aren’t becoming part of the training data for these models.
The people responsible for analyzing the results must be pulling their hair out wondering what kind of drugs their users are on with some of these prompts!
Speaking of training data, I thought the style of artwork was somehow familiar.
This one seems particularly reminiscent of Dave McKean’s art to me. Apropos since The Sandman debuts on Netflix Friday.
I wonder if they fed the model a steady diet of graphic novels.
Hmmm…
From Dave McKean’s biography:
In response to the development of artificial intelligence image creation programs like Midjourney, Dave created Prompt: Conversations with AI in twelve days. Through three short graphic story experiments, it explores the creative and ethical implications of this new technology, and asks how our working definitions of creativity, art and consciousness have shifted.
Maybe they captured his brain!!!
If you have a big mouth but a good heart, people will recognize your kindness. Even if you blabber on with your big mouth, as long as everyone knows that your heart is filled with kindness, they will love and accept you.
Please join our Discord server here, we are serious with the project.
Joined, very excited to start doing this all collaboratively in an organized way!
This here is the kanji for apologize. One of the trickier kanji for me. 謝
Mnemonic: You thought it would be a good idea to say the body measurements of everyone in your village and now you have to apologize.
When you do a bad thing you should apologize. And saying everyone’s private body measurements is definitely something you should not have done. So it’s apology time.
Not WK mnemonic related, but if someone would please provide this prompt to Midjourney, I’d be indebted:
“Dave McKean, held hostage and churning out art as fast as he can. He blinks three times if he needs help.”
(Aside: I just pulled out my stack of old Sandman and other Gaiman comics. I’d forgotten just how many illustrators were involved with that series over the years. Some amazing artwork. My kids grew up with Dave McKean artwork: The Day I Swapped My Dad for Two Goldfish, and The Wolves in the Walls.)
I did it and whilst I have no idea how those comics look like, I added , in the style of Dave McKean, to the prompt. This is what came out.
It knows what he looks like! I’m convinced it was trained on a steady diet of his artwork.
I’m going to have to find the book referenced in his biography. Sounds quite interesting.
Some examples of Dave McKean’s artwork. I love his work.
—-
Edit: a better link: https://www.davemckean.com/portfolio/childrens/
That’s the Children’s section of his online portfolio!
I hope your kids don’t have too many nightmares. 
Hahaha I thought about the same 
They came out perfectly weird. 
Coraline is the other “Kid’s story” collaboration between Gaiman and McKean. My kids loved both the book and the movie. That one is way scarier than the other two!
Thank you for the tip, I have to buy that asap. Unfortunately I didn’t know Dave Mc Kean but I like it a lot as well.
I don’t think kids would be naturally afraid of it, depending on the age of course. My son was scared the most of the movie ET. It was such a surprise, but after watching the intro, the boy walking to the garden finding ET he totally panicked. It was almost PTSD like and connected to a lot of hiding in the Oshi-ire for some weeks
Usually they don’t like the things I choose anyway, so I might buy it for myself.
Another try to give you an idea on how it works:
I enter “/imagine” then your text “Dave McKean, held hostage and churning out art as fast as he can. He blinks three times if he needs help.”
Here it would be possible to add some additional things, like seeds etc, but not a lot.
The result comes in 4 small image proposals:
Here you can do the following:
Enlarge on of these (with more details) or create another four variations of on of these.
Please choose what you want to do! (I will do according to your choices)
The reason why I post this is to show, that even if it is generated by an AI it is still the work of a human “artist” to a certain degree. You have to test with the input sentence and the image selection to get a result.
Not saying that it is comparable to an artist before AI, it is definitely easier to get results, but I think it is necessary to never forget, that an AI just cannot do certain things per definition.
That’s my personal view of the AI topic, feel free to condemn me
