How do you visualize an AI system training itself to build the ultimate warrior?
The Research Question
This question required more than visual design; it demanded understanding how machine learning actually works, then translating computational processes into narrative.
THE INVESTIGATION
Phase 1- AI Learning
Neural Networks
Decision Trees
Feedback Loops
Interaction Cycles
Phase 2 - Mapping Training Data
We created comprehensive flowcharts documenting every category of data required to train an AI warrior treating this as a research project in its own right:
MILITARY & COMBAT KNOWLEDGE Martial arts systems, historic warriors, military strategists, leadership philosophy, weapon-based techniques
ENVIRONMENTAL ADAPTATION Physical environment mapping, extreme weather combat, real-time sensory calibration
SYSTEMS & TECHNOLOGY Software architecture, hacking, data protection, quantum-speed management CULTURAL & HUMAN CONTEXT Music, humor, compassion, peace the full spectrum of human experience
Phase 2 - Mapping Training Data
We created comprehensive flowcharts documenting every category of data required to train an AI warrior treating this as a research project in its own right:
MILITARY & COMBAT KNOWLEDGE Martial arts systems, historic warriors, military strategists, leadership philosophy, weapon-based techniques
ENVIRONMENTAL ADAPTATION Physical environment mapping, extreme weather combat, real-time sensory calibration
SYSTEMS & TECHNOLOGY Software architecture, hacking, data protection, quantum-speed management CULTURAL & HUMAN CONTEXT Music, humor, compassion, peace the full spectrum of human experience
Phase 3 - The Building of the Ultimate Soldier:
Visualizations of the Learning Cycles
The AI begins as fragile code basic security protocols in a simple grid. Decision tree: warrior versus adversary. Immediate destruction.
A stronger core forms. The grid streams military history, Sun Tzu, Napoleon, and Alexander. Holographic warriors teach: Samurai precision, Spartan defense, Viking aggression. Martial arts data downloads.
Movements jerky. Failure. Death. Reset. Circuits glow brighter.
THE BATTLE ARENA
An armored core materializes. The grid becomes a dynamic battlefield, platforms shift, weather changes, terrain tilts. Six opponents attack. Decision trees branch with each choice, evaluating thousands of outcomes in microseconds.
Overwhelmed. Collapse. Reset.
ITERATION AS LEARNING
Fight. Death. Rebirth. Fight. Death. Rebirth. Faster and faster.
A thousand deaths. Each failure prunes the decision tree. Each rebirth optimizes performance. But incremental improvement isn't enough.
THE BREAKTHROUGH
The database expands beyond combat. All of human knowledge floods in: music, humor, compassion, peace, alongside violence, warfare, and tactics.
This holistic view unlocks potential that narrow training never could.
Thirty opponents. Every skill. Every tactic.
Victory. "Synchronization complete."
The insight: AI learns from diverse data, not just narrow optimization. The breakthrough came from receiving the full spectrum of human experience, mirroring how actual machine learning works.
WHAT THIS REPRESENTS IN MY PRACTICE
This project sits at the intersection of three modes of work:
RESEARCH
Understanding neural networks and machine learning well enough to visualize them accurately
DESIGN
Translating computational processes into a cinematic narrative that's emotionally resonant
PEDAGOGY
Exploring how learning happens in AI systems and human learners through iteration, failure, and holistic knowledge.