We discussed the idea of crowdsourcing / gamifying the process of finding interesting features in satellite imagery. The task seems well-suited to the approach since there is a lot of imagery and it can be difficult to identify features autonomously. The main problem we had with this idea was not knowing which features (or if there were any) would be scientifically useful to "farm out".