Can Artificial Intelligence be art? And if yes, can it be good?

Today on the Artalogue, Madison Beale sits down with artist Jake Elwes to discuss their art practice, as an early adopter of A.I., using it to challenge how we think about the world and technology through their artwork. Elwes does't separate art and tech, instead they use it as an innovative and generative medium. Elwes creates diffusion models that transform faces, words, and gestures into code. Bias, embedded into the datasets and diffusion models that are being used by almost everyone in our world, becomes something you can see when the work breaks down.
We discuss some of their works, like a diffusion‑driven interpretation of Sontag’s Against Interpretation. The Zizi Show, which has been the art work that introduced Madison to Elwes' work, “queers the dataset” by creating a unique data set with consenting and compensated drag performs, making moving images that transform into drag kings, queens, monsters and things that dance and lip-sync. The result is a cabaret that prompts us to think about bias, consent, and what it means to make art with tools that reflect us back in troubled but revealing ways.
Elwes and Beale talk about decolonizing data, art, tech and the problems that arise in the gaps between. We compare US and UK art education, unpack how normativity creeps into “perfect” generated images, and explore how far we can take Artificial Intelligence in the art world. The thread tying it all together is intention: tools are powerful, but human voice is the point. If you care about AI, art, drag, ethics, or how culture could absorb new technology without losing soul, this conversation offers a sharp, hopeful starting point for deeper thinking.
Follow Jake on Instagram @jakelwes and check out their website.

Connect with the Artalogue:

Madison Beale, Host
Be a guest on The Artalogue Podcast