FMA shows that market-based competition between autonomous agents can solve complex optimization problems without explicit fitness functions, suggesting economic dynamics may be a fundamental principle for discovering solutions in open-ended systems.
The Free-Market Algorithm is a new optimization method inspired by economics that doesn't need predefined fitness functions or fixed search spaces. Instead, it uses autonomous agents trading goods and competing in a market to discover solutions. It successfully found amino acids and nucleobases from raw atoms in chemistry, and predicted GDP with accuracy matching professional forecasters.