@Mo-Kanji I’m with you. This all sounds so cool but way above my skill level to implement, lol. I’ve followed the instructions but I’m obviously missing the next steps to actually create the pictures 
.
Heads up all we’ve now officially announced and opened the WK Mnemonic Images project and we’re going to collaborate on this with the goal of creating images for every single radical, kanji, and vocabulary in WaniKani. All 11,000 of them. We have a Discord server, an image database, a script to show them on WaniKani, and a bot for tracking submissions. Come join us! Learn more here:
Using DALL-E to make WaniKani images is genius! And it puts less pressure on me to draw them all.
I honestly got a bit burnt out from the pressure I put on myself to make so many drawings. lol 
I love this idea!
Yeah, the project in the post above yours to do it for all of the kanji seems pretty cool!
Just thinking … the whole time, have we been contributing to training these AIs?? … What have we done?? … Just kidding (sort of) love this thread 
This is the best discovery ever. OMG this is gonna change my game completely.
I recently saw some images pop up during my lessons of level 9 radicals and it was so cool!!
I’ve come to really not like this song after years of it being overplayed, lol
Lot of people with their backs to the camera there. AI gotta work on its repertoire a bit.
I don’t want to see his toe 
Well I, for one, am quite impressed at how accurately the AI produced a WaniKani statistics page.
Soon it can create whole fake accounts and go around the forums talking to us. Act as a spy for the budding Tofugu Empire.
Y-Yes… Soon…
How do I know it hasn’t already?
I wonder if people actually played around with NovelAI’s new Image Generation feature they added a few days ago. I decided to subscribe to the highest plan for $25 to play around with it since it allows unlimited picture generation.
I took the menomic for ride, 乗 with the following passage:
girl putting grain with nori in mouth and let it take a ride
This is what I got.
While the images look not bad for AI and gets the first part of the mnemoic somewhat correct, the AI can’t generate the image of a girl having a grain and nori in the mouth. I should probably try some other ones to see how it turns out in anime style.
I put in the mnemomic for older sister and this is what I got.
women in this city older sister with pet sheep
Played around with NovelAI a bit more at work, and I noticed that the last picture had a lot of inperfections. So, this is what I came up with instead. Instead of having sheep for older sister (姉） since the AI seems bad at generating them, I had a child sheep girl instead, and it works a bit better.
I came up with some other ones at work as well. Here is little sister (妹）:
Business （業）
Cat (猫)
Poor (貧)
Like (好)
Hate (嫌)
Translate (訳)
Princess (媛)
Sakura (桜）
Princess (妃）
Duplicate (複)
Comfort (案)
Okay, I have a bit too much fun with this. While this AI won’t replace actual anime art besides the limitations (honestly, actual fanart is still better), this AI is still useful to making kanji mnemoics in picture form and for brainstorming character design ideas as I do plan to comission some art for my anime blog later on. But I wonder if others give NovelAI a try with other WaniKani mnemoics since I am interested what other people will come up with.
Ah, just what we never knew we needed. All the WK kanji as anime girls.
Those are very cool. Just don’t look too closely at that football
I have finally created a dedicated repo for all the images I created so far. People can contibute to it and create a script to use these images on WaniKani. I know this will eventually blow up, eventually. Feel free to contribute to it.