AI x Design

01

Nektar – A Fragrance Concept Captured Through Gen AI

For this project, my goal was to create a product image for a conceptual perfume called Nektar, using MidJourney. I envisioned an artistic, modern aesthetic – not an outdoor scene, but rather a high-end studio composition that incorporated natural elements like moss and greenery, with controlled lighting and a shallow depth of field, as if captured with a 100mm f/1.8 lens.

Bringing this vision to life, I styled the set with organic textures and soft botanical details, all set against a clean white studio backdrop. To add a sense of organic presence to the composition, I introduced bees, who clearly showed an interest in the perfume. Finally, I used the Frame feature in Kling AI to animate the still image, turning it into a short, living visual.

MidJourney

Lowe – Exploring Consistency in Portraiture

This project explores the possibilities of consistent character rendering in AI-generated imagery. Through prompt engineering in MidJourney, I developed a recurring character — Lowe — and captured her in a series of realistic portraits. The intention was to create a figure both fairytale-like and fierce, the mythical mother of reptile creatures. Rooted in fantasy, rendered in realism.

02

Lowe – Consistency in
Portaiture

In this experiment, I explored the possibilities of consistent character rendering. Through prompt engineering in MidJourney 6.1, I developed a recurring character called Lowe, a mythical mother of reptilian creatures, brought to life through a series of realistic and emotionally raw visuals.

The concept naturally developed into a contemporary, fashion-forward narrative centered around Lowe and her loved ones in various poses and settings. She emerged with distinctive features; a model-like face, a fairytale aura and a striking sense of presence.


The visual direction reflects the character, drawing inspiration from Nordic aesthetics: cold, natural light combined with a muted palette of beige tones and natural fabrics.

AI x Design

01

Nektar – A Fragrance Concept Captured Through Gen AI

For this project, my goal was to create a product image for a conceptual perfume called Nektar, using MidJourney. I envisioned an artistic, modern aesthetic – not an outdoor scene, but rather a high-end studio composition that incorporated natural elements like moss and greenery, with controlled lighting and a shallow depth of field, as if captured with a 100mm f/1.8 lens.

Bringing this vision to life, I styled the set with organic textures and soft botanical details, all set against a clean white studio backdrop. To add a sense of organic presence to the composition, I introduced bees, who clearly showed an interest in the perfume. Finally, I used the Frame feature in Kling AI to animate the still image, turning it into a short, living visual.

02

Lowe – Consistency in Portaiture

In this experiment, I explored the possibilities of consistent character rendering. Through prompt engineering in MidJourney 6.1, I developed a recurring character called Lowe, a mythical mother of reptilian creatures, brought to life through a series of realistic and emotionally raw visuals.

The concept naturally developed into a contemporary, fashion-forward narrative centered around Lowe and her loved ones in various poses and settings. She emerged with distinctive features; a model-like face, a fairytale aura and a striking sense of presence.

The visual direction reflects the character, drawing inspiration from Nordic aesthetics: cold, natural light combined with a muted palette of beige tones and natural fabrics. ​

03

Interactive Neuroplasticity

In this work, I explore the intersection of creative coding and AI to generate interactive visuals and a conceptual experience. The animation, inspired by the work of Patt Vira, responds to the user’s mouse movements, enabling real-time influence over motion and flow. As the user guides the forms with their cursor, the system gradually self-organizes, revealing patterns and connections that echo the formation of new neural pathways.

Try for yourself.

Coming soon

Custom-built Gen AI model

I'm currently designing, building, and training a custom generative AI model from scratch using Python and PyTorch. The project is intended to deepen my understanding of artificial intelligence within the context of digital design — from model architecture and training data to the practical implementation of a fully functional generative system.

Generative AI explorations by Niklaz Hallberg © 2025