Diving Deeper into ARTificial Intelligence

AI art generated using text from Maya Angelou’s poem “Refusal” and “Dream” by Wombo.

According scientists, spontaneous artistic creativity is “one of the most mysterious forms of creative behavior, frequently described as occurring in an altered state of mind beyond conscious awareness or control.” Limb/Braun did a study on improvisation that linked this practice to parts of the brain used for language. These same researchers and Ron Eglash also looked at hip hop improvisation as an innovation in brain-to-brain connectivity.

The closest computer scientists and AI researchers have gotten to spontaneous creativity in machines, is through dreaming with neural network architectures (i.e., Deep Dream, generative adversarial networks or GANs). To this we can now add Natural Language Processing or NLP, which is concerned with giving computers the ability to understand text and spoken words. NLP is a form of AI that gives machines the ability to not just read, but to understand and interpret human language.

My chart showing various AI platforms and tools used to make art.

I recently created a chart to map out the different systems or areas of artificial intelligence that are linked to creativity (art). Then, I went deeper into new areas such as applications that combine GANs and NLPs, which sit in the middle as far as human-machine interaction and collaboration. My theory is that art, creativity and education is an effective way to interrogate the negative impacts of emerging technologies such as facial recognition AI.

Researchers say that,

Marginalized people often suffer the most harm from unintended consequences of new technologies.

Also, from a paper I recently found:

To address implicit (racial) bias and the lack of representation in AI fields, I proliferate social media with my AI artworks. Eventually, some people in those spaces took notice (ex. Smithsonian). I’m now exploring Dream by WOMBO, an app that lets users enter text prompts (up to 100 characters), pick art styles, add their own reference images, and turn different ideas into AI-generated paintings. Using text from a poem by Langston Hughes titled “Rivers” led to an ‘A-ha! moment’ for me:

I’ve known rivers ancient as the world and older than the flow of human blood in human veins.

AI art generated using text from Langston Hughes’s “Rivers” and “Dream” by Wombo.

I entered text from poems by Maya Angelou (also top image), Gwendolyn Brooks and I even compared/contrasted the poem “Ego Tripping” by Nikki Giovanni and “Ego Trippin’” by De La Soul.

It’s in the reach of my arms,

The span of my hips,

The stride of my step,

The curl of my lips.

I’m a woman

Phenomenally.

Phenomenal woman,

That’s me.

AI art generated using text from Maya Angelou’s “Phenomenal Woman” and “Dream” by Wombo.

I designed a pyramid so tough that a star

that only glows every one hundred years falls

into the center giving divine perfect light — Nikki Giovanni

Well, I’m da better brand, ’cause I’m a superman

I run the block with my circle, ’cause I’m nubian

I got the platinum rust, so don’t even fuss

’Cause DJ Paul, he’s down with us — De La Soul

Nikki Giovanni (left) and De La Soul (right)

We real cool. We

Left school. We

Lurk late. We

Strike straight. We

Sing sin. We

Thin gin. We

Jazz June. We

Die soon.

AI art generated using text from Gwendolyn Brook’s “We Real Cool”.

I became intrigued in how the tool matched/visualized words such as rivers and veins, pyramids and nubians, and Brook’s poem about teen delinquency. I was mesmerized by the way in which the software interpreted “span of my hips” and I started looking through my favorite poems for the words that I thought would create the most compelling images. Then, I thought that using machine learning and NLP tools would be a good way to bring in people who normally would not engage these technologies, as well as use the tools to teach young people about AI.

That’s me.