#ai #computing Humans success is often due to specialization. We must try to build hierarchy of specialized agents fulfilling simple tasks, the higher you climb the hierarchy, the sparser the goal. See [[Society of Minds]] ### Further reading - <https://thegradient.pub/the-promise-of-hierarchical-reinforcement-learning/> - <https://towardsdatascience.com/hierarchical-reinforcement-learning-a2cca9b76097>