The whole thing seems "simple" to me – graph of identities with url vetting/liking/approve-this-message-like actions, you don't need anything else. Reputation,

The whole thing seems "simple" to me – graph of identities with url vetting/liking/approve-this-message-like actions, you don't need anything else. Reputation, non-fakeness etc. can be derived from it for anybody - you just list identities you trust/follow (with weights?) and anything you look at can be scored. Virtual identities can also be created, ie. identity listing all links mentioned on HN (with positive sentiment only?), links from wikipedia etc. so people can follow those to create their reality graphs. The interesting part is that it doesn't claim universal truthness - depending on who you follow your results will be skewed towards their opinion of the world. Ie. if you follow MIT, Wikipedia and E. Musk you'll see different view of truthness than somebody following FOX News and Flat Earth Society for example. It could be interesting to focus on "dislike" marking (only?) as it may be much more lightweight to approach it from blacklisting side.