This is a project where simple AI agents (the purple cubes) learn to solve challenges by moving to colored blocks. Each generation, one agent is always the same as the best from the previous generation, while others learn from the best moves with small changes (mutations) to keep exploring new ideas. Occasionally, an anomaly happens where an agent starts over completely, allowing it to randomly discover a better strategy. With each round, the AI evolves and becomes smarter. Can you help guide the AI to find the perfect solution?
Want to know more on how it works? Ask me! Also turbo mode speeds it up slightly.