WaniKani mnemonics in image-form done by AI

That’s why there will be a job called prompt engineer in the future :smile:


Please hire me!
This is my dream job already.
I can “generate an image” and at the same time “learn Japanese” AND “knit”!
My favourite activities all in one :rofl:


The image in the lower left shows the most interesting premise to iterate on. I’m unsure if it’s wiser to “enlarge with more details” or create another four variations on the theme, but I’m suspecting the latter.

I also find it interesting that you’ve configured Japanese as your primary language for the interface. :smile:

What do the U1-U4 and V1-V4 parameters control?


I am blocked out from my computer now unfortunately but I will do it as soon as possible.

This is the result of making an upscale of the lower left (3) image

This is the result of variations of image 3

And these are variations of the upscaled image 3 from before.
I wanted to test if there is a difference in upscaling it first for making variations and there is!
It is possible that the part that starts to look like an eye in the “skirt” of image 3 could develop into an eye again if variations are made…

Please choose how you want to proceed.

In my impression it is a good idea to upscale first before doing variations to gradually add more details and to not upscale first to reduce details.

U is “upscale”. That makes one image the size of the four combined
V is “Variations”

A better way to change the image is to use another prompt

My way of “configuring” is to accept things as they happen :joy:
Because I am in Japan it turned out that way …


I’m not even at level 30 yet (level 9 currently), but I came across this kanji and checked out it’s mnemonic so that I could memorize it. It is so funny I can never forget it now. Can you guys do this one just for fun?

You put your fingers and an ear on a stool in the sun so you can get your photograph taken. This is a glamour photo, so you have to be in one of these weird, cheesy poses for it to come out jusssst right.


Imagine putting your fingers and your ear on a stool, now try to relax your head a little bit, and smile for the camera. This is a professional photograph, so you better do what you’re told!



This is perfect!


局 kanji mnemonicを下さいよ。

(I always associated it with the SCP Foundation. Wonder if they have invisibility cloaks?)


Apologies, but I appreciate it when people point out my mistakes in the language.

I think the よ here sounds very strange. It should just be 「◯を下さい」.

My Japanese isn’t good enough to know why it’s wrong exactly, but I’m pretty sure that it is. I think it’s probably because it’s a respectful word but still effectively an imperative, so adding an emphasizer feels unnatural.


Hello everyone - new to the forums but I’ve been watching this thread since it started and I just got access to DALL-E 2. Here are a few more:

泉 - The only creature powerful enough to withstand the force of these white water springs is the mythical centaur (せん). His horse body allows him to be stable and withstand the white water shooting up from the spring. Centaurs are pretty impressive creatures.

祭 - You have a tent with a jackhammer in it for the festival! Festivals need booths where people can do activities. In this tent, festival goers can use a jackhammer and make a lot of noise, perfect for a festival! This festival is for cyborgs (さい) only. Picture the cyborg you made getting excited that you’re taking it to the festival. That’s why all of the activities have to do with heavy machinery. The rides are mostly just cyborgs letting other people ride them, but it’s still pretty fun.


Sorry, in the case you didn’t notice it, I added the new images to my old post which is maybe confusing…


Your nation’s flag sits on top of a caped mouth. This is the symbol for the “Bureau of Secret Things.” It’s a special department of your government made to keep things hidden.
level 8 局



Aesthetically, I actually like the upscale “image 3” version with what most looks like an eye at the bottom. Pretty interesting image to me, even with only a very tenuous relationship to the prompt!


It looks really interesting. In terms of workflow at this point it would be better to change the input to something that could possibly create a result closer to your initial idea.

Personally I think AI will never substitute any human. Maybe to create one interesting image, but a Manga would be soon starting to freak out and become too bizarre to be interesting. Maybe, maybe I am wrong :sweat_smile:

This is the maximum upscale:

The red dash became a women!