Tonight I created 700 gorgeous illustrations in my sleep and this WILL change the creative industry #ai

This is an excerpt from my upcoming ai art related presentation

This title is no clickbait. Sorry to disappoint the skeptics. I did create 700 illustrations and I'll share them with you at the end of this article. But first things first.

Today you can choose your favorite ai tool from a variety of different ones and use them to generate digital images. The range of tools is wide. #DALL-E #Midjourney #DreamStudio #NightCafe #Croiyon offer acceptable to good UX and exciting results. 

The famous dalle loading bar ;)

However, creating several images in parallel is only possible in a somewhat restricted way. At the same time, this approach (prompt " 1 min wait " prompt 1 min wait) consumes a lot of time and is not always ideal for a successful operation in a corporate context. 

At some point, I was bothered by the constant short waits and even though they can be mitigated by working in multiple tabs, I was in need of a better solution. This arrived in the form of Stable Diffusion, which is an open source alternative to the above. It can also be used on your home computer - as long as you have a non-apple system with a graphics card and plenty of RAM. Unfortunately, I do not belong to this group. 

Instead, I was able to get Stable Diffusion running on a Google Colab Pro server, and adapted it to run multiple jobs in the background. Everything is controlled via the Google Colab web interface. All generated images are stored in a folder of my choice via a Google Drive connection. 

Thus, I can order my Generations in the evening, sift through the results in the morning and select the best variations. And that is what I did.

Future Usecases

Certainly, many creatives will integrate AI into their everyday work in the future. But the real advantage lies in letting the systems continue to work overnight. Here we can fully utilize the power of these systems and pick out the best pices in the morning. 

In this case, it was these 7 graphics out of 700 for me. However, I will not really use them.  

So what did we learn:

✅ These systems are powerful but upping the output requires taking control into your own hands.

✅ The ability to create visual content while sleeping will change, how creatives and designers handle tasks of visual work in the future.

✅ If I promise 700 images, I will deliver. Here you can download all 700 for free. Feel free to use them as you please. If you do you can credit me, but it's not mandatory. BTW. DISCLAIMER: You can find characters in some of the generations. This was not prompted. Looks like I have to figure out, how to get rid of them in future production cycles.

❌ Currently I cannot export high resolutions images with this solution. If some of you have found a way to get high res exports with google colab, let me know.

If you want to know why stock photography will die soon, check out this article

👉 Follow me -> Jacques Alomo