As the compute demands for machine learning and artificial intelligence applications continue to grow, co-design techniques and neuromorphic hardware have been touted as potential solutions. New emerging devices like memristors, atomic switches, etc have shown tremendous potential to replace CMOS-based circuits but have been hindered by multiple challenges with respect to device variability and scalability. The time is ideal for a significant re-think of neuromorphic hardware design. In this paper we will use a Description-Design framework to analyze past successes, under-stand current problems and identify solutions. Engineering systems with these emerging devices might require the modification of both the type of descriptions of learning that we will design for, and the design methodologies we employ in order to realize these new descriptions. We will explore the advantages and challenges of complexity engineering over traditional approaches to neurmorphic design, the various changes that will accompany it and offer a possible path forward. Success will represent a radical shift in now hardware is designed and pave the way for new paradigm.