December 13, 2023

Drawn at Fusion in Albuquerque:

TABLET DRAWING

Drawn on a Galaxy S9+ Tablet using Sketchbook:

 Sketchbook

 Sketchbook

EzGif

Animated drawing progress

 Sketchbook

ModelScope

EzGif

The animated drawing progress

on AI steroids

 Sketchbook

The tablet drawing

 Sketchbook

EzGif

 Sketchbook

ModelScope

EzGif

Version #1

Robot

 Sketchbook

ModelScope

EzGif

Version #2

Dancing Robot

Leonardo.AI added a "Live Canvas" feature on December 1st, which updates the drawing in real time time, nano-seconds after one makes a mark or changes the model:

Leonardo.AI "Live Canvas"

So I uploaded my tablet drawing into Leonardo.AI's Realtime Canvas, selected the "painting" AI model, and gradually pushed up the "Creativity Strength" settings, saving the image at every nudge.  

Leonardo.AI "Live Canvas"

The settings in "Realtime Canvas"

Then I made an animated GIF of the progress, using higher and higher "Creativity Strength" settings:

Leonardo.AI "Realtime Canvas"

EzGif

Using the "painting" model in Leaonardo.AI,

I gradually increased the "Creativity Strength"

AI ART

The above second tablet drawing (the non-animated version) was altered in Leonardo.AI with a model based on my long drawings:

Leonardo.AI

Leonardo.AI converted the top drawing

into the bottom drawing

Leonardo.AI

prompt: "Sculpture of  a woman sitting on the floor

 in the style of Bellini KrrrlLong"

Leonardo.AI

prompt: "Sculpture of  a woman sitting on the floor

 in the style of Bellini KrrrlLong"

• Does Midjourney "train" on the images I generate in their Discord?  The next step of AI is to "train on synthetic" data, generated by AI...but will that mess up the results?