Imagine a new AI Art system that is so powerful and game-changing that nearly EVERY known AI platform...NightCafe, Accomplice, ArtBreeder, and more...ALL adopt this system as a new standard! This AI Art system, known as Stable Diffusion, upped the AI Art game SO MUCH that previously popular platforms, like Disco Diffusion and VQGAN+CLIP, are little more than vague memories. Even the fan favorite, MidJourney, is adapting Stable Diffusion into its platform! And here's the kicker...this transformation took place in just 2 weeks...ALL of it!
So, what does that mean for you and me? For starters, our ability to create amazing, mind-blowing art of nearly ANY type and genre just exploded by a factor of 10! Here is another exciting benefit now available for you and me...because of the differences in how these various platforms work, the results we get from each will vary! It's kind of like the old axiom: “A rising tide lifts all boats!”
In addition to Deep Diving into the various iterations of Stable Diffusion, we will also Deep Dive into OpenAI's Dall-E 2, Stable Diffusion's chief competitor. We will discover what's similar and what is different. For instance, Dall-E 2 offers an InPainting option that is simply MAGIC. InPainting is when you erase part of an AI-generated image to allow the AI to paint in a new variation of the erased section. This feature can be used to create some amazing art outputs that no other current program can perform. The Circus Illustration above is an example.
Lastly, I have been experimenting with OpenAI's text generator, GPT3, to generate prompts for use with your favorite AI Art Generators. I have found the outputs to be beyond anything I could typically imagine, like having the AI explain the emotion and characteristics of a Van Gogh painting! In other words...using AI-Generated Prompts to create AI-Generated Art! And, as always, I will include TONS of demos, case studies, insider insights and much more!