Efficient progress in science is only possible if researchers can access reliable research findings and data in a manner that renders experiments reproducible. Access to research output has to be equitable, and the community should consider whether it is beneficial if publishing is solely controlled by a small number of commercial entities. At the same time, the community also has to be prepared to invest in infrastructure and processes that ensure research findings are stably archived and amenable to interrogation both by human and machine. This includes next generation search and discovery tools and deep interlinking. Scientific output is exploding and more than ever this necessitates the filtering for reliable and useful information. Mechanisms built on peer review remain the most effective means to achieve this goal. It will be argued that OA is an important driver to ensure broad access to the scientific literature, but that we must not drive down costs so much as to erode the quality control function of the publishing process. It will also be argued that there are multiple steps between the bench and a review article to share research findings. These steps synergize to accelerate scientific progress without eroding quality control and reproducibility. If framed correctly, Open Science will preserve quality assurance, while also allowing the sharing, archiving and discovery of all significant research outputs