Ladies and Gentlemen today we are going witness an A.I. fight between OpenAI vs Stability AI
Folks both contenders are heavy weights of the undisputed Nvidia A100 Tensor Core GPU and also both are equipped with the impressive deep generative neural network diffusion model.
Fighting for Stability AI is Stable Diffusion and fighting for OpenAI is DALL.E 2 and both are trained with unique fighting abilities.
Stable Diffusion is open source and opponents are well aware this trained fighters autoencoder that can denoise white noise to land knock out realistic image generation punches at 768x768.
Here is the Stable Diffusion python package.
https://pypi.org/project/stable-diffusion-sdkit/
DALL.E 2 is closed sourced yet rumors are that this well pre trained fighters rise to fame is because of their decoder of white noise that connects knock out abstract image generation punches at 512x512.
Here is the DALL.E 2 python package.
https://pypi.org/project/dalle2-pytorch/
This fight will have five rounds refereed by thy greatness myself and each round is determined who lands the best knock out text to image generation.
Round one fight
"Wooden slat exterior on two story house with windows"
Stable Diffusion:
DALL.2 E:
First round goes to Stable Diffusion.
Round two fight
"Automated Factory"
Stable Diffusion:
DALL.E 2:
Second round goes to DALL.E 2
Round three fight
"Teslas with mountains in background"
Stable Diffusion:
DALL.E.2:
Third round goes to Stable Diffusion.
Round four fight
"Auckland City CBD"
Stable Diffusion:
DALL.E 2:
Fourth round goes to DALL.E 2
Round five fight
"Beach with sunset"
Stable Diffusion:
DALL.E 2:
Fifth round goes to Stable Diffusion.
And Stable Diffusion is the winner by unanimous points count Ladies and Gentlemen.
Enjoy 😊 Alec Bellamy