In 2007 and 2008, I submitted proposals to NSF and NIH to extend the infrastructure we had built for Project Implicit into [1] a [Virtual Laboratory][2] for improving rigor, transparency, and sharing of research what we referred to as an "open source science framework", and [2] an [international participant pool][3] that would be a community-based infrastructure for behavioral researchers to collaborate on sampling. Both had an underlying them of improving research rigor and reproducibility. Neither were funded. After multiple attempts, the ideas were put in stasis. The latter idea of a community-based infrastructure lost some of its value proposition as commercial services have emerged providing sampling solutions for behavior researchers, such as Mechanical Turk. The former idea kept percolating and was revived in early 2011 when Jeff Spies was looking for a dissertation topic and kept returning to our shared interest in improving research practices. His technical background and understanding of scholarly communication made it conceivable to pursue the OSF as a dissertation project. He did and we supported the infrastructure initially with proceeds from speaking fees from my implicit bias lectures. The first broader use of OSF was to help coordinate and manage the Reproducibility Project: Psychology with its hundreds of collaborators. That project and the OSF received some press attention in 2012 that resulted in a few private funders to reach out about potentially supporting the work. The Arnold Foundation made a big bet on us leading to the founding of the Center for Open Science in January 2013. [1]: https://osf.io/3v2wy/ [2]: https://osf.io/784rn/ [3]: https://osf.io/3v2wy/