





Women in GenAI National Leadership Council awarded Drue Kataoka the 2025 Artificial Intelligence Innovator Award at the inaugural Women in Gen AI summit and awards hosted by Fenwick & West in Silicon Valley. Watch the video here.

Women in GenAI National Leadership Council awarded Drue Kataoka the 2025 Artificial Intelligence Innovator Award at the inaugural Women in Gen AI summit and awards hosted by Fenwick & West in Silicon Valley. Watch the video here.

How does an AI model perceive a person? How does it begin to understand who they are? Why is it that the fragmentary and elusive nature of human memory makes it more beautiful?These questions drive some of my latest experiments—training custom AI models on my own image, then placing them in worlds I imagined and animated.Walking through a mirrored desert of my mind, several AI versions of me exist in a space that is both familiar and surreal, a construct of memory, data, and imagination. A reflection of self, refracted through technology.The appearance & disappearance of each figure down to the fraction of a second and the meaning behind that is deliberate. There are always easter eggs & layers upon layers in my works. Sunglasses by Christian Dior Couture. Watch the full video here.

💡 Is “AI art” more art or more programming? 🤔
Watch the video here
💡 For this artwork, I trained a custom LoRA. 🎨🤖 LoRA stands for Low-Rank Adaptation and is a technique for fine-tuning large machine learning models, creating a bespoke adaptation.
💡 So I trained the LoRA on a curated dataset of my Ambrosia series, optimizing the model to retain the intricate textures and color dynamics 🎭 of the original physical works. I further refined and expanded the digital model with RunwayML for generative augmentation and Topaz for upscaling and noise reduction, preserving fine details while allowing for creative extrapolation. 🖌️✨
💡 You know, I think the beauty of these tools lies in their combinatorial power—each iteration opens up infinite new pathways for artistic evolution. 🚀🔮
💡 By the way, the original Ambrosia series is meticulously fabricated with real-world materials, presenting unique engineering challenges. 🏗️ Ambrosia has become one of our signature works, renowned for its dynamic interplay with light—an interactive medium we’ve developed with over 169 billion physical configuration permutations. 💎💡 Conceptually inspired by neural networks, the Ambrosia mimics the way information propagates through weighted layers—just as light refracts, reflects, and transforms through their intricate structures, continually modifying, obfuscating, and revealing new forms. 🔄🔬
💡 So, taking images of physical Ambrosia works, encoding their intricate structures into LoRAs, and transforming them through a custom AI workflow feels like coming full circle—bridging the physical and digital realms in an endless feedback loop of creation and iteration. 🔄🌍💡

Watch me speak 7 languages in 1 minute. I fabricated an avatar (using various AI tools & techniques) that I can use to say anything in any language. I made the original videos and images in my art studio in Silicon Valley where I was standing in front of one of my real life Ambrosia artworks. What do you think?
Watch it here.