Stable Diffusion: A new kind of art is coming...

Greetings,

Today I'm going to share with you a new technology that is appearing in the world eventually that could change drastically what art is, or even more this could be a new kind of art. What if you could write something and a computer create a new picture for you from zero? Well, there are new AI technologies in BETA version that is doing exactly this!

Many of them are currently closed and you can use it only if you have an invitation, but I found one that is open source, it is called Stable Diffusion, the bad part is that is require a lot of computing power so you will need a good PC to run it. I as a developer could not resist and wanted to try it.

The first step was download the source code contained in the official github repo:

• https://github.com/cmdr2/stable-diffusion-ui

You will get a ZIP file, then unZIP it and move the folder that you got to your C:/ disk at the root.

There you will have to run the .cmd file what will start the project (this could take a lot of time and is needed ~20GB in disk)

And if everything goes fine you will get the UI where you can start generating images.

It is the interface, you can write anything you want and it will create for you images that will be for your own, it means the copyright is yours and you can use it as your own draw. Let's see my first image:

As you can see I wrote A warrior elephant that can fly breaking a peace of ADN and I got the previous one, I got to reduce the GPU usage cause to generate a very single pic it is taking a lot of time for me.

Let's see another examples:

But it is not too easy, you need to be as much specific as you can or you will get bad results like this:

If you want to test it I would recommend the following:

• 8 GB RAM (16 GB recommended)
• 50 GB SSD as free space.
• A dedicated GPU.

And now, do you think this could replace artist? or it is a new area to explode?