Autonomous Agent Simulations
and Other Experiments

Sudoku Agent Swarm

Autonomous agents solving Sudoku by reducing stress. Each number is an agent that moves to reduce their own stress rom not being unique. 

Sudoku Agent Swarm 

Tony Robbins 6 Human Needs

Simulation of autonomous agents collectively finding global maxima. Each agent is unique, having in their set of needs: Certainty, Variety, Connection/Love, Significance, Growth and Contribution.

Six Basic Human Needs - Agent simulation 

Nested Three Body Problem

Exploration lab to see whether nested three body systems can exhibit balance between order (attractor states), and chaos.

Nested Three Body Problem 

PicBreeder style CPPN Playground

Inspired by Kenneth Stanley and his Picbreeder. Not really agents but vastly interesting anyway. Hidden neurons use a mix of transfer functions: sin, cos, gauss, id, sigmoid. Original Picbreeder uses CPPN‑NEAT with innovation-number crossover. This demo uses a fixed feedforward topology and performs uniform crossover over parameters.

PicBreeder style CPPN Playground 

Video2Sound via FFT & CPPN

Visual complexity evolved into sound via Fast Fourir Transformation and CPPN neural network that iss evolved by the user.

Video2Sound via FFT & CPPN 

Healing Sound Breeder

Various Sound is generated by ANN's. The user selects the most relaxing ones. The preferred sound-ANN's combined to new audio-generating ANN's. Gradually the sound can become very relaxing.

Healing Sound Breeder 

Autonomous agent sorting

Inspired by Michael Levin and his work. Agents

Agents sort themselves witout global supervision or control. The user can place obstacles making it harder for the agents to sort themselves.

Agent Sorting 

Self-Organizing Criticality - falling cubes

This simulation visualizes a simple physical system that can exhibit Self-Organizing Criticality (SOC).  The characteristic of self-organizing critical system is that the system continuously organizes itself near a threshold between stability and instability without external tuning.

Self-organizing criticality - falling cubes 

Z - displacing pixel in images 

The perspective is everything. This experiment lets you view a 2D image in 3D as from a fourth dimension. Experiment is inspired by one of the chapters in my novel: "Dö for dig - En roman om AI".

Z - displacing image pixels in 3D 

Fish Ecosystem simulation

Plankton grows continuously. Small fish eat plankton and survive better in schools. Large fish hunt small fish. Reproduction happens after enough food is consumed.

Fish Ecosystem simulation