As in previous sections we group results by the initial objectives.
**1. Science Funding:** We provide a flexible model of the evolution of sciences that serves as a building block in a funders objective by telling her (a) how scientists choose questions, (b) how much they invest into those questions, and (c) how they inspire future researchers. We show that a combination of a static cost friction and a dynamic externality lead to suboptimal pace if scientists are purely funded by cost reductions. Instead, ex-post rewards such as prizes are a key element to incentivize beyond statically optimal, inspiring research. However, they work best in combination with cost reductions. We show that having _occasional_ moonshots (*not marsshot*) are best to inspire generations of "normal scientists". We have also obtained first results on providing feedback to scientists, showing that, when costly, it is best to give feedback on early, and when the researcher had had a fundamental misunderstanding. With established scientists, feedback has less effect on decision making and therefore can be left out.
**2. Publishing Null Results:** We can show that the failure to recognize non-findings censors the belief of researchers and leads decision makers to become pro-active when prudence would be best. However, they often do not hinder the progress of sciences, as researchers would shy away from opening up endeavors that previous generations have failed on despite trying hard. It turns out that to overcome these blockades (which is beneficial for society) additional incentives potentially from funding institutions are needed.
** 3. Competition and Dynamics in the Market for Ideas:** Our first results here indicate that when future demand of the outcome of an idea is uncertain and there is a need to potentially refine ideas later on, there is a strong first-mover advantage in which the first-mover takes a moderate step leaving following competitors the avant-garde market.
