Replacing the current react-and-regulate approach to managing the complex outcomes of technological change demands new institutional arrangements for the production of knowledge and technology. Collingridge’s famous dilemma tells us that in the early stages of technology development not enough is known about potential impacts to choose the most beneficial development paths – and that once impacts do become apparent, it’s too late to shift to better paths because of technological lock-in and vested interests. The dilemma assumes a linearity of technology development: first we innovate, then we respond. This view leads to the current story of social conflict and regulatory gridlock. Yet linearity is not inherent in technology development, but rather it reflects institutional arrangements in which technology development and technology governance are largely separate activities – organizationally and temporally. These arrangements are neither inevitable nor immutable. Once technology is understood to be a product of continual choices made by humans in politically mediated settings throughout the innovation process, the governance problem can be recognized as not one of prediction and control, but rather one of awareness, openness, and course-correction. New institutional structures can be designed accordingly. At the Center for Nanotechnology in Society and the Consortium for Science, Policy, and Outcomes, both at Arizona State University, we have been developing and testing theories, methods, tools, and institutional arrangements for making the implicit choice processes of research and development more explicit, inclusive and transparent. We describe the activities themselves as “Real-Time Technology Assessment” and the cumulative effects of these efforts as “Anticipatory Governance.” The goal is to improve the potential for desired social outcomes of technological advance, and reduce the likelihood of undesired outcomes.