Experimental Way

Transforming organizations through interdisciplinary design

Ai Soroush Doustvandi

Personal Collection

SixtyEditions

Timeframe

2019 2022

Services

AI-Driven Products, Branding, Engineering

New Approach...

For years, long before tools like Midjourney and ChatGPT entered the spotlight, I pursued an experimental approach to AI-driven design by leveraging Python and various coding frameworks. This early, self-directed research involved crafting custom scripts and algorithms that interfaced with design systems, generating unique forms, patterns, and structural concepts. By exploring computational geometry, parametric modeling, and data-driven creativity, I was able to push design boundaries and translate abstract ideas into tangible prototypes.

 

With the rise of Midjourney and other advanced AI platforms, I seamlessly integrated these newer tools into my established workflow, allowing for even more rapid and fluid iteration. This blending of homegrown code and modern AI services has expanded my design repertoire, delivering an agile, technology-enhanced approach. Whether developing architectural concepts, product sketches, or experimental art pieces, each creation now benefits from an ongoing dialogue between algorithmic experimentation and cutting-edge AI innovation.

Table lamp Rocky Style Soroush Doustvandi
tango7

Results

Designing lighting products with AI introduces both unprecedented opportunities and distinct challenges. On one hand, generative models and algorithmic processes enable rapid exploration of novel forms, improved light distribution, and unexpected design directions. On the other hand, it requires careful data preparation, model training, and iterative refinement to ensure the AI’s outputs align with real-world engineering and manufacturing constraints. This integration demands a blend of technical acumen and creative vision, ensuring each design remains both aesthetically compelling and functionally sound.

 

Recognizing AI’s growing potential, I decided to increase my investment in these advanced techniques and develop a custom-trained model tailored to lighting design. By feeding it precise information—from photometric data to materials engineering parameters—I can guide the AI to produce unique solutions that strike the right balance between art and science. While it’s undoubtedly a challenging path, requiring diligent experimentation and constant optimization, the payoff lies in uncovering groundbreaking product concepts that push the boundaries of contemporary lighting design.