HyperArt (2025)
HyperArt is both an artistic project and a conceptual framework that reimagines creativity through collaboration between humans and artificial intelligence.
The idea was inspired by HyperText, a concept so embedded in our daily digital lives that we hardly notice it anymore. Reading Computer Lib/Dream Machines, a visionary book from the early days of personal computing, reminded me how radical these non-linear, interactive systems once were, opening up new ways of thinking and creating. Today, with AI, I feel we are living through a similar pioneering moment. Just as personal computing once expanded creative possibilities, artificial intelligence is now transforming my own artistic process into a network of interconnected stages, where human and machine contributions continuously feed into each other.
Borrowing this spirit, HyperArt proposes a dynamic, non-linear workflow where an artwork evolves across multiple iterations. Each step, whether human-made or AI-generated, feeds back into the whole, creating something greater than the sum of its parts.
With HyperArt, the focus shifts from final output to the ongoing exchange between human creativity and AI, questioning and expanding what it means to create in this new era.
Moreover, the network of inter-relationships between one stage of the artwork and the other ones becomes itself an artwork: some networks are symmetric and harmonious, while others are more chaotic and unorganized.
So far, I’ve developed two main examples:
Self-PortrAIt
Starting from a selfie, I used Stable Diffusion, ControlNet, and IPAdapter to generate AI portraits styled after my past works and later merging them into an evolving animated video.
Hybrid Painting
Beginning with a hand-drawn sketch, I generated an AI reference to create an oil painting, which has been later digitized and animated. The painting can also act as a trigger to an Augmented Reality experience showing the animation. I’m currently experimenting with other variations of this process and having fun creating new ways of merging machine and human creativity.
Sara Sisti: Who I am
Since 2019, I’ve worked across various companies developing text mining systems, computer vision applications, and recommender systems, applying state-of-the-art AI techniques. In 2024, I shifted to freelancing, focusing on projects that combine coding, AI, generative AI and data analysis, with a strong preference for projects involving the creative usage of these tools. I particularly enjoy playing with open-source tools like ComfyUI, Automatic1111, and Stable Diffusion.
Alongside my technical work, I’m deeply involved in artistic experimentation. I love merging traditional and digital media such as painting and clay with digital art, creative coding, generative AI, and TouchDesigner to create hybrid, surreal compositions. One of my experimental pieces was featured in the CVPR Online Art Gallery 2024 & 2025.
My favorite creative approach draws from surrealism, both in theme and process. I’m fascinated by the subconscious, allowing free associations of images and concepts to guide my works. Through augmented reality, generative transformations, and iterative reinterpretation across media, I explore ideas of consciousness, perception, dreams, and the hidden complexity of the everyday.
Academically, I’m also active in research and teaching. I’m a research fellow at IULM University, contributing to PRIN 2022 “ALTEREGO”, a project investigating how software surrogates can simulate intentionality and awareness. I also teach a Masterclass on Stable Diffusion, focused on creative AI workflows using open-source tools; a module on Digital Trends in the Master’s program in Fashion Communication and Luxury Strategies at IULM University.
For me, creativity is a space of total freedom, where technology and art can amplify each other. If you’d like to explore my projects, you can visit https://www.sarasisti.com/.
Or you can follow me on instagram @sarasisti.art @sarasixti, twitter/X @sarasixti,