Hacker News Reader

Is there a way to loopbin user feedback on summarization that improves the algo? How could you incentivize users in the process?

I guess we would have to use Machine Learning to incorporate that.