Convergence: Collaborating Across Realities

Convergence is a collaborative project born out of a fascinating journey that melded the traditional paintings of acclaimed Contemporary Southwestern artist Lawrence Lee with modern artificial intelligence technology.

I got to know Lawrence through our work together on establishing an industry-wide convention for secondary sale royalties in the cryptoart space. Lawrence, who has painted thousands of quality "magic people" or Shamans over a half-century career, provided an incredible foundation for a new kind of creative process.

The Process

I approached Lawrence about a collaboration, knowing he had a lifetime of artwork to build on. We realized that by combining his immense catalog with my somewhat unique artificial intelligence art techniques, we could create something completely unprecedented.

Using over 250 of his original Shaman paintings, I trained several Generative Adversarial Network (GAN) models. Because early AI struggled to build good-looking torsos and faces simultaneously, I built different models—one trained on half-length subjects, another specifically for headshots, and a third that produced surrealistic, otherworldly blobs.

The goal was to expand the creative palette like a hallucinatory dream. I directed the AI to produce a googleplex of possible random outcomes, introducing noise and artifacts to throw the model off just a bit and create strange, unexpected elements.

Lawrence and I then hand-selected the best outputs. From there, Lawrence used his digital painting skills on an iPad using Procreate to unearth these new shamans, enhancing them to pull out the best features and release their powers. Sometimes he literally invented an ancient god out of an abstract AI-generated blob.

The Results

This iterative feedback loop—where I would inject Lawrence's corrected, digitally enhanced works back into the training set to teach the AI model all over again—created an evolving, surreal lineage of new beings.

The resulting series, Convergence, was released as weekly drops on SuperRare. It stands as a testament to what is possible when you bridge a lifetime of traditional easel mastery with the unpredictable artifacts of machine learning.

Read the original editorial piece covering the project on SuperRare.