This paper seeks to show problems which follow from wrong assumptions about the key drivers of innovation policy and theory. A theoretical evolving network model is proposed for that purpose, taking into consideration innovation opportunities that are not necessarily driven by technological discoveries, R&D or knowledge diffusion. Innovation is defined by a new combination of means of production, and is introduced to the initial condition by rewiring a regular graph model. Dynamics of the diffusion process are captured by the statistical reconstruction of the solutions for recurrence equations within a discrete time perspective. Simulation results show that innovation gives rise to essentially unstable turbulent dynamics. Stability analysis implies an alternative interpretation of the nature of innovations and their dynamics. As a source of economic change, innovations are conjectured to be dependent on scale and diversity. They are closely linked to novelty, rather than to previous states of knowledge or technology.