The effectiveness of deep learning for specific tasks is driving innovation in hardware, which may lead to the discovery of intelligence systems that operate differently or on a larger scale than deep learning. However, adding dendrites to neural models can solve problems in continuous and rapid learning similar to biological systems.