Mosaic world title

Mosaic world is a 2-D virtual landscape of colour in which artificial life agents evolve to ‘eat’ the colour of surfaces. This is because in this artificial world – as in the real world – the virtual life ‘critters’ need energy to survive, though in their world they only get this energy by ‘eating’ colour – some being good and some being bad.

Our research has yielded a number of important findings about the nature of colour vision, how and why it evolved the way it has, and even about the emergence of multicellular organisms.

The strength of Mosaic World is the ability to relate the resulting behaviour of the ‘brains’ of artifical life agents directly back to their ecological history, and in doing so answer the questions about the computational principles required for robust visual behaviour.

Go to Downloads here for our Mosaic World demos.

Related published papers

Evolving visually guided agents More
Evolving multicellular artificial life agents More