I am a professional graphic designer, primarily involved in creating layouts for advertisements, posters, presentations, and websites. I also teach IT (Digital Skills) and I like to experiment with graphics and technology, especially emerging technologies like AI. I'm really interested in exploring how a machine perceives both our natural world and our unnatural world (art, culture, imagination).

I'm currently on day 555 of my Duolingo streak, so I've got to know those characters quite well. I was inspired by some posts on Bored Panda where creators had reimagined Simpsons characters or other cartoon characters as real people. The Duolingo characters are very basic and I wasn't sure if they would translate well into real people (plus some of the Simpsons ones I've seen are borderline scary), so I thought, what if I took it halfway... Fleshed out (so to speak) but still cartoony. The most obvious example of which is the Pixar movies.

I wouldn't say I have perfectly captured the "Pixar style" but it was used as inspiration, although I appreciate consistency across the different images isn't 100% by any stretch of the imagination.

Picking a favorite image is like asking me to choose a favorite child! Haha kidding. Oscar was the one I started with and that was really good fun. He's very expressive and playful in the Duolingo app and I wanted to try to capture some of his personality in the image. So that's why he is in a Hispanic setting, with some colorful bunting as a nod to him celebrating his LGBTQ+ identity (although purposefully not a rainbow) and drinking from an oversized wine glass like one of the animations in the app. 

I started creating these images by compiling the different elements of the scene I wanted to create. So let's take Oscar as an example. I researched images of Mexican streets and downloaded a selection. Downloaded some wine glasses. And of course a picture of Oscar. These were imported into Autodesk Sketchbook (an Android app similar to Procreate or Photoshop) and a rough draft of the image was made.

This was exported and uploaded to OpenArt.ai, which is a great AI website that allows you to use their basic tools for free (you can collect daily credits from following them on Discord). With the rough draft uploaded into the "image to image" setting, I then have to craft the "prompt". This involves describing the character as I see them, from their appearance to their mood, and what they are doing. I also have to describe the style I want.

Initially, you get so very bizarre pictures coming out, so you have to refine the prompt and keep generating until you get something decent. This was the final prompt: "A still from a 3d movie of Disney overweight Mexican man with a mustache, pink shirt, green pants, drinking wine, 3d blender render, Pixar inspired, Disney, clear detailed beautiful face, dark eyes, adorable. Summer vibes. 8k octane renders unreal engine, surrealism". Then the picture is upscaled and exported. Put back into Sketchbook for some touchups and alterations, and that's it.

In my professional life, I enjoy being about to use my skills to communicate ideas and opportunities to people through the work that I produce. With the AI and digital art side of things, it is about learning and evolving my skills to influence my professional work and have fun. I've enjoyed how it can also bring a little joy to people, although the backlash from the anti-AI protesters that infects a lot of online communities does tarnish that somewhat.

I understand and appreciate their viewpoint, but equally, I don't agree with the aggressive and rude way that they approach it. Hopefully, as the world gets used to AI, it will calm down a bit and AI and real art can live happily side by side. Who knows?!