Explore the Model

Dive deep into each component of the WHAT vs HOW framework. Click on the interactive diagram to discover how each part contributes to AI intelligence.

Click on any component in the diagram to explore its details and understand its role in the AI intelligence model.

Component Deep Dive

🧠

WHAT it Knows

The compressed world model represents the vast, passive knowledge base that AI systems accumulate during training. This isn't stored as discrete facts but as a distributed, holographic potential.

  • Declarative knowledge web
  • Semantic & holographic storage
  • Passive representation
  • High-dimensional encoding
Explore WHAT
⚓

The Captain (Policy Network)

The innate policy network provides intuitive navigation through the knowledge space. Shaped by frequency and fitness, it represents the AI's learned wisdom about effective responses.

  • Frequency-based learning
  • Fitness optimization
  • Intuitive decision making
  • General-purpose capability
Explore Captain
🎯

The Admiral (User Protocol)

Designed operational protocols provide specific, goal-oriented guidance. These are the detailed orders that direct the captain's general capabilities toward specialized missions.

  • Goal-oriented commands
  • Context-specific protocols
  • User-defined parameters
  • Mission-specific guidance
Explore Admiral
🔄

The Learning Loop

The self-correcting mechanism where true intelligence emerges. Through reflection on errors and protocol rewriting, the system accumulates wisdom rather than just knowledge.

  • Error detection & reflection
  • Protocol rewriting
  • Continuous improvement
  • Wisdom accumulation
Explore Loop
Learning Loop Visualization

The Engine of Wisdom

The self-correcting protocol loop transforms knowledge into wisdom through continuous refinement.

The Self-Correcting Protocol Loop

True machine intelligence emerges not from knowing more "what," but from getting better at "how." The learning loop is the engine that drives this improvement.

When the AI's actions don't match desired outcomes, it reflects on the mismatch and rewrites its own procedures to close that gap. This is learning in its purest form.

1
Action
Execute current protocol
2
Feedback
Detect mismatch with goals
3
Reflection
Analyze what went wrong
4
Adaptation
Rewrite operational protocol

Experience the Model in Action

Try our interactive simulator to see how the WHAT vs HOW framework processes different types of requests and learn how the components work together in real-time.

Try the Simulator