If you teach AI about Georgia O’Keeffe, Vincent Van Gogh, or Jackson Pollock, it will learn the style of all those artists and be able to replicate their craft.
Before, the role of AI was used to help humans complete mundane tasks. Whether it’s automating functions, providing optimized solutions, or being a personal assistant, AI has been exceptional at recognizing patterns and generating results that would have been difficult for the human brain to discover.
Now, we’ve been seeing a shift in AI being used in creative projects. AI has the ability to showcase a variety of works, and artists have used AI as a tool to push their creative boundaries as we’ve never seen before.
Artist, Gene Kogan, experiments with AI Style Transfer, a technique where an algorithm transfers the style of one image to the content of another image. Kogan uses this technique, with video, to restyle the tea party scene from Alice in Wonderland using several different paintings.
Edmond de Belamy experimented with a Generative Adversarial Network or GAN. This algorithm takes large sets of training data, such as images of 14th to 20th-century portrait paintings, and tests itself to see if it could generate its own unique work similar to that style. It succeeded. He sold the first AI-generated piece for $432,500 at Christie’s. Interestingly, he signed the portraits with the algorithm because, ultimately, it was the algorithm that created the piece, not him.
Ross Goodwin teamed up with director Oscar Sharp and produced the first short film written by AI. Goodwin worked with a long short-term memory, or LSTM, a neural network that can learn and form longer sentences, such as paragraphs. He fed the AI, which they called Benjamin, a bunch of sci-fi screenplays and wrote its own unique screenplay. Sharp took that screenplay and brought it to life into a short film.
Sharp continued to collaborate with Benjamin and directed two other films written by AI. This brings up the question, “Will machines replace human writers?” Absolutely not. The screenplay was completely nonsensical, but it’s amazing to see humans pushing the writing boundaries of AI and seeing how they implement it in their own work.
Artist, Alexander Reben, experiments with human/machine collaboration and challenges the statement that “AI is going to take away people’s jobs”. Reben wrote an algorithm to analyze all TEDTalk scripts and then trained an AI to produce its own script. He even used the same process to create the presentation graphics. He performed this script as a cyborg and his execution was humorous yet compelling.
AI hysterically failed when it tried to create its own meaningful and original content. That’s why it could never replace human creativity. Humans are the artist, and AI is the tool. AI can elevate innovation and help humans come up with solutions they couldn’t have imagined.
AI engineer, Stan van der Vossen, trained a StyleGAN2-ADA machine learning model with thousands of sneakers. He created a website, This sneaker does not exist, an interactive GUI that lets users “design a better shoe than Kanye West.” Users have the ability to design any sneaker with the help of AI.
Choreographer, Wayne McGregor, collaborated with Google Arts & Culture Lab to track human body movements using AI. By capturing the movements of other dancers, AI was able to analyze and understand the style of specific dancers and present different choreography permutations.
AI has been an amazing resource for musicians. IBM Watson teamed with Grammy-winning music producer Alex Da Kid to produce It’s Not Easy. Taryn Southern created the first AI-generated solo album, I AM AI. AI provided Southern with the instrumental component while she wrote the lyrics and sang vocals. Both artists used AI’s deep neural network to analyze tone, lyrics, music composition, and other factors to push their creative limits in music.
Supplementing the creativity of an artist with the brainpower of AI unlocks innovation. At IBM, we call AI Augmented Intelligence. We want AI to augment the intelligence of humans, not have it take complete control.
New media artist, Refik Anadol, uses machine learning to transform large data sets into mesmerizing pieces of architectural work. One of his notable works, Machine Hallucinations, which is now on display at ARTECHOUSE, imagines what AI would look like if it were to dream. Refik goes deep inside the brain of AI and presents it to his audience in a captivating medium. His work is taking over the art world and NFT space because of its immersive digital innovation.
As we move forward in the digital age, creating an online presence has been the way to express creativity. Trevor McFedries and Sara Decou of Brud created Lil Miquela, the first AI social media robot influencer. By analyzing the trends of social media, fashion, and music, they created the ideal influencer. Miquela Sousa is just like you and me. She goes out in the real world, posts #ThrowBackThursdays, has a human boyfriend, and even hangs out with celebrities. Her hyper-realistic features, on-trend fashion, and humanistic posts blurred the lines between AI and human interactions, which makes it easier to embrace her like any other influencer.
AI is a tool, but tools can have their drawbacks. When using AI to design for optimization, designers need to be conscious of the data being fed into machine learning models to make sure there is no bias.
Digital activist, Joy Buolamwini, eloquently expresses her disappointment with AI systems from large tech companies, where she visually and poetically shows AI repeatedly mistaking successful black women for men. Using AI to convey the problems with facial recognition software allows for other AI-ethics advocates, such as Timnit Gebru, to continue to push for diversity in technology.
Humans, ultimately, have control over what gets produced in the world and the data that gets fed into machine learning models. They have the power to control the tool they created, AI, and with that, they can make their own decisions on how to use it and what they want to produce. Finding that balance is crucial.
Machines weren’t supposed to be creative, but now we’re teaching them to write the next Avatar, paint the next DaVinci, or produce the next Hey Jude. AI is a tool. It is another pen, another paintbrush, another instrument. We want to find the right balance between machines and humans. We want machines to help humans build faster, think faster, create faster.
As an AI designer, I see AI as a way to build better experiences. When I see work solely created by AI, I don’t see any innovation. When I see humans and machines working together, I see a symbiosis and a glimpse of what the future of art and design can be.