Virtual Autonomous Learner

This page offers key points related to the virtual autonomous learner.

For a gentle introduction to the concept check out the post: The Virtual Autonomous Learner: An Introduction.

If you are curious how I came up with this idea you probobley should read this post: The inception of the Virtual Autonomous Learner.

The Virtual Autonomous Learner (VAL) is a 3D agent currently in development.

  • Virtual – The 3D agent lives in a 3D computer environment that uses a dynamic physics engine to simulation real world physics.
  • Autonomous – The agent is not controlled by anyone. Its behavior is its own. Its actions depend on its internal state and its environment.
  • Learner – Like a child, the VAL agent has many developmental stages and has to learn every thing from scratch.

In short, VAL is a computational cognitive construct that uses artificial neural networks, based on biologically plausible cognitive models and architectures, to generate human-like behaviour. VAL’s construct is a framework that utilizes and manipulates different cognitive architectures. VAL also, uses a rapid prototyping process that promotes the development of new computational models in an attempt to find the ideal cognitive “blank slate”. And, finally, VAL takes an ecological perspective to facilitate the agent’s developmental and learning processes, this approach builds a construct on an embodied cognition foundation.