What is generative art?
What is “generative art”? In a fairly recent definition by Jason Bailey from Artnome magazine, “generative art is art programmed using a computer that intentionally introduces randomness as part of its creation process.” So, why is generative art relevant for deep dreamers? AI art based on generative adversarial networks, GANs, is a subcategory of generative art. An aside: The Deep Dream Generator does not use GANs, it is based on convolutional neural networks.
Selected visual generative artists
Let me present a few inspirational generative artists below. Some of them have created online programs or apps that will let you create your own art, even for use with DDG. Dive in!
Artist and PhD student Memo Akten focuses especially on machine learning and expressive human/machine interaction. To better grasp this concept, watch one of the scores of generative art videos he has posted on vimeo, such as “body paint.”
If you’re interested in Akten’s philosophical approach to machine learning, take a look at his website, read an inteview with him in Artnome magazine, or take a look at his GitHub profile.
Like all the new-media artists who are at home in programming, Joshua Davis has a presence on all key social media. He explains how to use programming to generate artwork – that is, above all to write software to make art – in skillshare classes and on YouTube.
Perhaps of special interest to deep dreamers is that Davis developed an interactive generative art app for iPhones called Reflect as early as in 2009. This app allows you to swipe to create patterns chosen by randomization algorithms.
Davis’s generative art method has been dubbed “dynamic abstraction.” The artist has taught at the New York School of Visual Arts, speaks at conferences and conducts workshops, e.g. at the MoMa in NYC.
Deep dreaming features among the techniques photographer, digital media artist and instructor Adam Ferriss applies to his works.
The digital and print publication Parallax Collab, which explores cutting-edge links between art and science, provides a comprehensive biography of the artist. There, you will find an explanation of Ferriss’s methods and sample eye-popping creative coding video/gif material.
Woud you like to see more? Look up Ferriss on Instagram. Also, his GitHub page contains numerous scripts, tutorials and even a deep dream program.
Benjamin Fry is a data visualization expert. Together with Casey Reas (see below), he has developed Processing. This software is an open-source programming language and integrated development environment for the electronic arts and visual design communities. The source code is available on GitHub.
Fry created the generative image shown here entitled “Ben Fry” with the dreamlines drawing machine.
As a code artist at the Google Cultural Institute, Mario Klingemann ‘(“Quasimondo“) works with Deep Learning and large image datasets to make stunning digital art. Klingemann is a pioneer in the area of neural networks, computer learning and AI art. Browse his gallery of neural art (“Memories of Passersby”) created with GANs at Onkaos.
Klingemann has uploaded images of his early art – indicentally, alongside amazing galleries of open-source images e.g. from the British Library – to his flickr website. His starry night image, incidentally, may be shared and adapted.
Raven Kwok is a visual artist, animator and creative programmer who uses computer algorithms and software processes to generate videos.
Try out his Noise Turbulence Doodles, funny little doodles that you can generate by dragging your mouse across a blank screen.
Kwok has posted code-based generative lyrics videos on vimeo and stills from the videos on flickr.
John Maeda is one of the pioneers among generative artists. Maeda straddles the line between many fields. In his own words, “I began as an engineer, then moved to art / design, and then to research / leading / operating, and then into business / tech / investing.”
Maeda is sought after as a speaker and has authored numerous books about computational design and leadership.
In the late ’90s Maeda recruited several brilliant artists/technologists into MIT’s Media Lab to help work on “Design by Numbers,” including Ben Fry and Casey Reas. Fry and Reas took Maeda’s “Design by Numbers” into classrooms around the world and eventually built their own free platform called Processing that could be shared outside of universities and used by anyone with an interest in learning to sketch with code.
Manolo Gamboa Naon
Manolo Gamboa Naon (Manoloide) is a creative coder and generative designer. See his mesmerizing videos on vimeo, visit his projects on GitHub, look at his work e.g. on Instagram, Behance, Dribbble or Twitter.
Naon’s generative art stands out as invariably “pleasing to the eye.” In an interview with Naon in Artnome magazine, Jason Bailley acclaims Naon as the prodigy of generative art.
MIT graduate Casy Reas, who teaches at UCLA, hosts a gallery of his work on his own database website (active until 2016). More recent material is on his homepage. You can also find links to his social media presence, code platforms, print publications and video material on the website. Reas initiated Processing with Ben Fry (see above) as early as in 2001.
Moreover, Reas authored a standard programming textbook for the visual arts with Ben Fry. The aim of this book is to teach visual artists software literacy. Download a free digital copy here.
Jeremy Rotzstain, a Canadian coder artist, developed a photography-based painting app for iPhones, PhotoRibbon, as well as an iPhone art app called Swipin’ Safari. While these apps may no longer be available, Rotsztain’s art is.
Follow Rotsztain on flickr or vimeo or check out his page on GitHub.
Leonardo Solaas is a philosopher, artist and programmer from Argentina. To get links to his work on flickr, Twitter, Vimeo, Instagram and Behance, just go to his website.
The image shown left is a “not painting” Solaas programmed in Processing.
Solaas programmed an amazing generative art maker, the dreamlines drawing machine: The script switches randomly over eight sets of formulas that transform the color values of a pixel into angle and velocity values.
I tried it out. See three stages of an ever-evolving “birthday cake” dream below. Using the machine is simple – enter words into the subject bar on the website. I chose “birthday cake” as the subject I wanted the machine to dream. Next, the machine performed an image search on the web. Finally, the results fed into the generative process and constantly transformed a dynamic drawing.
It’s amazing to watch an image appear and steadily metamorphose. Who knows, you may just generate images to use for your own work on DDG.
Would you like to find out about other computational artists?
One resource is this blog post on the online educational platform Kadenze. Moreover, the Generative Art Wiki is an excellent source of material on generative art; it also provides links to numerous generative artists’ websites.