Golden: Mapping human knowledge

Hopefully if there's ever a global fallout we'll have the entirety of human knowledge store deep in the ground for future generations lol. This project is amazing though.

To angry self: I dont entirely trust Wikipedia to give an objective/apolitical view of things all the time: there've been times where articles have been noticably influenced by a defacto, anachronistic political (progressive) position. I wouldnt mind having a redundant source of information where the possibility exists of current politics corrupting its articles' integrity

so this is a for profit company that intends to use free labor contributions to exploit search engine SEO rankings and subsequently attract useless millions of daily visits and then you exit it for hundreds of millions a few years from now. It worked before, why wouldn't it work now? there are always enough dumb people who will help you make a fortune for free if you understand tech capitalism and manage to fool them; good luck anyway.

For a tool that has the high goal of "mapping human knowledge" and use A.I., I'm just not very impressed by the lack of APIs and no mention of a data model.

I feel that to map human knowledge and make it accessible and understandable for machines, the data model should be put front and center. For me that means a knowledge graph (preferably with SPARQL/OWL capabilities).

Were's the metadata? The wiki pages don't even have schema.org data! I'd be more interesting if golden provided UI improvements over the querying and presentation of data from Wikidata and/or DBPedia, or if golden's editor would make it easier to annotate the content of the wiki pages with RDF data.

This is absolutely fantastic. I'd been distraught by Wikipedia policy of "notability" and deleting valuable articles that volunteers created by pouring several hours. I firmly believe that no knowledge, no human is small enough not to be "notable". I also like increased focus on tooling which Wikipedia has failed to deliver in all these years. I think Wikipedia was good start as trying to emulate encyclopedias of 18th century but in new age we need to move on to AI-first knowledge graph that can have billions of nodes where each node in the graph could be anything from some human to some object in my backyard to entire textbook.

There would obviously the question of how do you prevent misinformation and falsehood. If you want to scale to billions of nodes, moderators aren't going to cut it. One possibility is leveraging community and what I'd call chain of trust. For example, community can flag, upvote, downvote. This doesn't result in deletion but simply a signal to the reader about how trustworthy content this may be. The chain of trust mechanism can improve this further by inferring contributors that users have trusted previously. The StackOverflow like gamification for contributors can create wonders here. In addition, you can allowing users to create their social network so they can build their personal chain of trust. Another possibility is to put untrusted articles in draft domain and move them to main domain as trust level is increased. The key is to avoid deletion of content and retain it somehow so it can be improved and evolved.

Now the things I don't like about Golden:

When signing up, it forces bio to 140 chars. Why? Why not collect more knowledge about authors? Not artificially limiting information should be the point here, right?

I also find current interface very cluttered and unfriendly. After signup I was greeted with topic of blockchain and cell based meat occupying most of my screen real estate. I don't care about either and half-visible conversations under each topic does not help. How about asking me what I'm expert in? What are my interest? Add some algo magic to recommend topics for contribution?

I also don't like UX at all. For example, this is page on Bitcoin: https://golden.com/wiki/Bitcoin. The menu that suddenly breaks after quick intro hurts my eyes. The typography is straigning. The left menu just hard to grasp. On page for cluster, you get giant list of contributors on right which I care less. You can say whatever about Wikipedia but they got all these stuff right.