Hmm. I would work on a scientific community that makes it easy to publish Science-to-Public books (to increase public interest in science), to get peers to review your paper easily (to decouple it from journal) and help people know and recongnised the talented scientist in their area. It bothers me that the science being done every day by clever people doesn't get the attention and funding that it should.
Would a collaborative filtering engine that was able to discern inputs/feedback that were reliable versus not be useful to your endeavor? That's a key piece of what I'm working on: https://qht.co/item?id=8086694
