Smart electronics have developed ubiquitously to assist people in everything from navigation to health monitoring. The rise of complex electronics relied on rational design of platforms to build ever larger and more complex circuit networks and for frameworks to test those electronics. Biochemical circuits have also seen dramatic advancement in the last two decades within the field of DNA nanotechnology. As with electronics, DNA nanotechnology applied rational design to DNA molecules to build ever more complex biochemical networks that, beyond current electronics, also retain a significant measure of biological compatibility and plasticity akin to many networks of biological origin. Well situated for promising applications in diagnostics and therapeutics, advancing DNA nanotechnology devices will also rely upon larger platforms and testing frameworks. In roughly the last decade, researchers have been building upon the invention of DNA origami, a technique allowing the robust construction of biomolecular nano-structures capable of precise nanometer positioning of proteins, nanoparticles, and other molecules. DNA circuits have computed on the nanostructures; DNA robots have moved nanoparticles, made choices, and have even sorted cargo on the surface of a nanostructure. The complexity of circuits and devices continues to rise. In this thesis, we will discuss our contributions to the field of DNA nanotechnology by developing design rules and systematic approaches to controlling nanostructure complex assembly. These rules and approaches allow for the construction of molecular structures with a tunable diversity, large systems approaching the size of bacteria yet retaining nanometer precision, and biological plasticity inspired dynamic systems for arbitrary reconfiguration. Using a DNA origami tile tailored for array formation with a high continuous surface area, we create a framework inspired from molecular stochasticity for programming DNA array formation and gaining control over diversity of global properties through simple local rules. Three general forms of planar networks, random loops, mazes, and trees, were manipulated on the micron scale upon the self-assembled DNA arrays. We demonstrate control of several properties of the networks, such as branching rules, growth directions, the proximity between adjacent networks, and size distributions. The large diversity, in principle, allows for a wide, but tunable, testing environment for molecular circuits. By further applying these principles to subunits of finite assemblies, variable components may be mixed with fixed components potentially opening additional applications in high throughput device or drug screening. Next we turned to expanding the platform size biochemical circuits may be built upon. While DNA origami allows nanometer precise placement, the size remains roughly below 0.05 um2. Toward making large arbitrarily complex structures with only a set of simple tiles, multi-stage self-assembly has been explored in theory and for small DNA tiles. None were successful experimentally with DNA origami. We developed a strategy for DNA origami: a simple rule set applied recursively in each stage of a hierarchical self-assembly process, and to significantly reduce costs, a constant set of unique DNA strands regardless of size. We also developed a software tool to automatically compile a designed surface pattern into experimental protocols. We experimentally demonstrated DNA origami arrays approaching the size of small bacteria, 0.5 um2, with several arbitrary patterns, each consisting of 8,704 specifically chosen pixel locations with nanometer precision, including a bacteria sized portrait of a bacteria. The large platform opens the door to more advanced molecular circuits for applications such as diagnostics. Finally we demonstrated control over the dynamics of DNA origami reconfiguration in tile arrays. In an approach we call DNA tile displacement, we showed that a DNA origami array may have tiles arbitrarily replaced by another tile, including tiles of another shape or surface pattern. We also demonstrated control over the kinetics of tile displacement and performed several general purpose reconfigurations of DNA nanostructures. Examples include sequential reconfiguration, competitive reconfiguration, cooperative reconfiguration, and finally the scalability of multi-step reconfiguration as demonstrated through a fully playable nano-scale biomolecular tic-tac-toe game. The major ramifications are a plasticity more common to biology than to electronics—molecular platforms with arbitrary patterning that can reconfigure an arbitrary part of the nanostructure in an arbitrary order based on environmental signals. In principle, such reconfiguration can allow advanced circuits with the capacity to adapt to environmental needs or heal damaged components.