Connectionism: Modeling the mind with neural networks — LessWrong