Facebook, Twitter and TikTok feeds. Google results and ads. Retailer interactions. Government services. Traffic lights. Algorithms control—or at least greatly influence—our daily lives in a huge number of ways. The use of algorithms is certainly not new, but corporations and governments rely on them more and more to anticipate our personal decisions and manipulate how and when we consume information and ideas.
There have been countless examples of AI and algorithms gone awry. From the criminal justice algorithm PATTERN over-predicting recidivism among inmates of color to AI tools introducing bias into the hiring process, we’ve all seen the horror stories in the headlines. You may be feeling like SkyNet can’t be far off.
But it’s not all bad news. There are also a lot of interesting things coming out of the worlds of AI and algorithms, especially in the realm of art.
Creating art with algorithms
I’ve started playing with algorithms and images via DeepArt, which uses algorithms to combine images and art styles. It’s a really fun, accessible way to experience some of the magic and wonder of what an algorithm can do. Here, I combined a photo of myself and my wonderful pit, Playa, with the style from one of my illustrations. Got a few minutes? Try creating some art for yourself. While DeepArt is now defunct, you can use Prisma for a similar experience.
If you want to go even deeper, AIArtists.org is a really interesting site for learning about and experiencing AI-created art. The site shares work from top AI artists, provides information on the history of AI art, and encourages visitors to create their own work.
Creating Art with Algorithms – Original Images
Creating Art with Algorithms – Result
Algorithms and music
Musicians and programmers Damien Riehl and Noah Rubin have created a music-copywriting algorithm designed to stop the never-ending list of lawsuits that occur when someone hears a hit and is convinced that the artist stole their idea. As there is a finite way to arrange musical notes and beats, the duo decided to copyright every available 8-note, 12-beat melody combination and offer them for public use.
The algorithm generates 300,000 unique melodies per second, then outputs MIDI files to a hard drive, which automatically copyrights the melodies. From there, Riehl and Rubin waive the rights to their interests using Creative Commons Zero, which places the work in the public domain.
As it turns out, I find the artistic possibilities of algorithms to be far more inspiring, and I hope you take time to explore them further. And best of luck as you traverse the murky waters of our algorithmic reality.