Getting economy-based systems (e.g. prediction market, DAO, insurance) right is extremely difficult. How to reward/penalize users to prevent exploits and ensure resultant behavior is as expected?
Today, teams resort to theoretical analyses based on strong assumptions about user objectives and rationality. That approach, while insightful, is very time-consuming and potentially very unreliable.
The Incentivai approach is to simulate a large number of scenarios. Bad actors are modeled using ML agents who can explore the system in search of possible exploits and clever ways of maximizing their profits. Good actors' objectives model relevant aspects of human behavior: risk aversion, speculation, imperfect information.
The Incentivai tool allows designers to observe how the performance of their system changes as they vary assumptions about human behavior, discover new possible exploits and iteratively improve the design.
We'll keep you updated when the best new remote jobs pop up.