Use the combination of a neural network and a genetic algorithm to build a creature that can do any task.

The Evolution simulator is a joint project of the MIT Center for Advanced Virtuality and the MIT Media Lab.

The goal of the application is to build a machine that can perform any task. It doesn't have to be a specific task, it can be anything. The idea is that the AI will learn to do it and improve over time.

A neural network is a type of computer that is used to make calculations. In this case, the calculations are done by the network of neural cells.

A genetic algorithm is a type of computer algorithm that uses a combination of rules and instructions to find the best solution for a certain problem.

In this case, the problem is to build a creature that can do any task.