Dooders is an open-source research project focused on the
Project description
Dooders
Reality works; simulate it.
Overview
Dooders is an open-source research project focused on the development of artificial intelligent agents in a simulated reality. The project aims to enable the conditions and mechanisms for cognitive agents to evolve and emerge in a digital environment.
A Dooder is an agent object in the simulation with an amount of causal control. It acts in the simulation only as long as it consumes Energy.
Take a look on how a Dooder will learn and act, get an overview of the core components of the library, or read why I started the project.
I will also be documenting experiments in substack. Including the results from my first experiment.
The code, content, and concepts will change over time as I explore different ideas.
Everything in this repository should be considered unfinished and a work-in-progress
How to use it
from dooders import Experiment
experiment = Experiment()
experiment.simulate()
experiment.experiment_summary()
# Example output using the default settings
# This simulation ended after 53 cycle when
# all Dooders died from starvation
{'SimulationID': 'XGZBhzoc8juERXpZjLZMPR',
'Timestamp': '2023-03-09, 18:20:33',
'CycleCount': 53,
'TotalEnergy': 634,
'ConsumedEnergy': 41,
'StartingDooderCount': 10,
'EndingDooderCount': 0,
'ElapsedSeconds': 0,
'AverageAge': 14}
For more details, see the Quick Start guide.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.