After seeing negotiation without systematic fairness, a developer created Mediator.ai, combining Nash bargaining theory with LLM capabilities. The platform addresses a fundamental problem: people struggle to define utility functions for complex decisions, while LLMs can compare different agreements to estimate preferences accurately.

The tool works by having each party privately state preferences through LLM interviews. These become a fitness function for a genetic algorithm that iteratively refines potential agreements. In a case study, bakery co-owners Maya and Daniel used it to resolve their $80k investment dispute over ownership percentages.

Mediator.ai produced a 60/40 split that neither had proposed, with Daniel able to restore his 10% share by returning full-time for six months or forgoing $24k in distributions. The agreement included a management salary and shotgun buy-sell clause, creating a resolution balancing past contributions with future expectations.