I just realized that we could use a prediction market to decide in which open-source common-good projects to spend time on, considering the expected rate of success and impact of each.