My AI Learning Journey

Let’s face it. Using machine learning for creative works nowadays is becoming more common place and easier to do. Using closed-source AI (OpenAI, Gemini etc.) can only get you so far and you basically get what you get (or what you pay for). Since the advent of Open-Source models and interface options like Comfyui, us plebs can now run, refine & control our image generations on our own hardware for the price of the electric bill.

Here is a “journal” of some of my experiences learning Comfyui and running it local on my 3060ti. I will add to this journal as my experiments continue…

BACKSTORY

When generating art with image and video models, it’s crucial to have a plan. Without one, you can quickly…

STEP 1: Testing the waters

The first step is to gather some sketches and “play around” a bit. ComfyUI stood out as having the…

STEP 2: Try some more…

Original Artwork This is a “dripper”. I will not go into detail on their origin, or why they are…

STEP 3: How Basic Can I Go?

I wanted to see how much I can get using my simplest “collage sketch” ideas as a basis for…

Stay tuned for more…

I am currently looking for work and as such I have limited time to work on this project. Abe’s Journey is not unlike my own and I hope to complete both one day!

My plan for the next steps of this story are:

Finalize a style for the artwork, characters etc.
Create a LoRA (most likely multiple ones) for characters
Finalize form of story (Comic book, Children’s book, animated comic etc…)
Write the Story (it’s currently in a bunch of random notes and doodles)
use final product to sell the concept to an animation studio with the intention of cartoon series.

Thanks for stopping by and having a look! Feel free to drop me a line if you have any comments or questions 😀

This is the original artwork I used to generate the tall concept image. I’m still amazed at how basic you can be to guide the final generated image.