Convergence Analysis of Machine Learning Algorithms for the Numerical Solution of Mean Field Control and Games: I - The Ergodic Case