It's remarkable to see how Zooniverse has branched out into so many fields and different projects now. An interesting andecdote I have from grad school is that another student was working on a project, a component of which involved classifying galaxies. He was very interested in developing a machine learning approach, but his advisor suggested instead of spending several months refining the approach and tuning the tools available, to put it on Zooniverse instead. The dedicated folks at Zooniverse finished the entire sample of more than 1e7 galaxies in less than twenty four hours. It was straight forward to set up their program and the information they received was very high quality. His takeaway was that (for how much of a bandwagon machine learning is at the moment in astronomy) it's unique that there are a huge number of dedicated individuals donating their time for the betterment of human knowledge in the Zooniverse program, and it often may not make sense to painstakingly apply the hottest ML techniques when something like it is available.