Home Creators Posts Import Register Favorites Logout

Content

Hey all!

As I get started on Aurora, I'm looking to get some new artwork for it. The elemental mandala worked well for BTDEM, but I'm looking to get new art, and a new image, to help make Aurora visually distinct from BTDEM.

Also, the license on the artwork is super restrictive (I was brand new, and trying to do a friend a favor), and the artist refuses to budge a hair on the license, or expand it in any way. Like, I offered them $1k to expand it, but they refused.

Part of me wants to try and get a professional artist, the rest of me would love something from the community!

I have a $400 prize pool for it (which, if the winner is sufficiently pursuasive, could go up). $300 for first place, and $100 to be spread out among a sufficient number of 'high effort' runner ups. If there's enough strong contenders that the prize pool is getting small, I'll significantly up it - I don't want someone to put in a ton of work and get nothing for it. Basically, I'm trying to set myself a reasonable budget, but I know in my heart of hearts it's futile. I am IN FAVOR of using software and programs to help. For example, https://arcanigen.thatrobhuman.com/ could probably get you almost all of the way there if you spend the time tinkering with it.

Required: The eight base elements are present. They connect to the 36 advanced elements. Ideally, the shape is a magic circle or magic-circle like. Or just really pretty.

For those that need a refresher on the elements and combination:

https://docs.google.com/spreadsheets/d/1ED3DISd3Tm4finJh7DFg3DpN1BubricpfOGVSHdysFQ/edit?usp=sharing

The contest will run for approximately three weeks. Submissions can be made here or on Discord. I might leave the window open a little longer

Cheers!

=====(Silly hat on)

In other news!

I've heard that some of you like litRPG. That you need more DINGS! in your life. That you crave the sweet, sweet hit of dopamine as numbers go up. Fear not! After extensively studying the market and the genre, I have crafted the perfect story. Just look at these reviews!

"A wonderfully fantastical story from the very first Ding! all the way to the last.  The MC goes through a complicated series of character progression and levels up many times over the course of the story, starting at a low level and eventually reaching a much higher level.  My only real complaint is that there are a few places where there could have been some more Dings.

Ding!"

"This is truly an avant garde piece of the LitRPG genre with a distilled bluebox 'Ding!' injection that every chapter delivers, after some, unfortunately necessary, worldbuilding and plot, at the beginning.This MC power levels at pace I've never seen before, with no meaningless side characters, side plots, or  any sort of unlikely or controversial romance getting in the way! Just pure levelling goodness after they get lucky, or are blessed by the heavens, as some might say.Surely this MC shall ascend and be recognized as the greatest MC of all time by anyone that reads this apocalyptic tale! Ding! Ding! Ding!"

"Absolutely perfect for anyone unsatisfied by the incredible paucity of skill level-ups in modern litRPG, an excellent throwback to some of the older style and telling the full story of someone at the very bottom progressing all the way to the very top with nothing but a single advantage and the willingness to endure tremendous trials. It really immerses you in his journey, making you feel every victory and trial he undergoes on his way to godhood."

"Selkie!" You might be asking yourself "Where can I read such a masterpiece!?"Fear not! It's free on Royal Road, right here: https://www.royalroad.com/fiction/198552/the-dingpocalypse Note: RR insisted it get the AI-generated tag, even though it's algorithmically generated. Fair enough. Every line of code was carefully hand-crafted by me, and I did the cover personally.

Rate! Review! Get the joke to the top spot on RR!

Also, I've only posted about 20% of it so far, need to finish posting it.

Cheers!Selkie

Files

Comments

Sam Friend

A question from the uninitiated - what does Algorithmically Generated mean?

Arusalan

What is Aurora?

SelkieMyth

AI is a bit of a 'black box'. "We trained software that this is a picture of a bird. This is a picture of a fish. This is a picture of a bike", and it thought quite hard and went "okay, I think I know what a fish is". But we don't know what's going on inside the AI's head. It took millions of dollars to train, and there's a lot of ??? going on. Algorithms are much older (oh god this hurts to say) and much more simple. "Pick A. Add one. Write the message "Ding! Leveled up A, it's now level X+1". Record the new number. Repeat the process." There are clear, strictly defined steps, and you know exactly what the computer is doing and where.

Sam Friend

Ahhh gotcha. Makes sense.

SelkieMyth

WIP title of my next project! A return to Pallos!

Aoinfinity

It’s mildly encouraging that they made you add the AI-generated tag, because that means maybe someone on RR will force people to tag the dozens of obviously AI-generated stories that *don’t* have tags. …I mean, I don’t actually expect it, but it would be nice, you know?

bcdp

An often cited example (not sure if made up or not) of "what's going on inside the AI's head"is this story: Researchers trained an AI system to distinguish photographs of tanks from photographs without tanks. The system appeared to perform very well on its training and test data, but it had actually learned to distinguish sunny conditions from cloudy conditions rather than recognizing tanks, because all photos with tanks were taking on the same day, a different one as the photos without tanks. This is a classic example of bad training data. The AI grouping for the training data was good. It was just not what was intended, because the training data contained a strong correlation with an irrelevant side effect. That could have been easily avoided by using more diverse photos. Another example: A study showed that the tested image classification AI had a bias towards texture over shape and other cues. So trained to classify animal images, when presented with a cat with overlayed Elephant skin: Where humans would see a cat with strange skin, the AI strongly guessed at an Elephant. Neither is wrong, since it has both Elements, but it shows that an AI doesn't necessarily learn the same markers we look for, not following our intuition.