3 ms·
In between. I have some half-baked ideas but not enough time to implement them. But in a nutshell the idea is to do a pagerank-type calculation to compute peo
by lisper 7y ago
In between. I have some half-baked ideas but not enough time to implement them. But in a nutshell the idea is to do a pagerank-type calculation to compute people's reputations so that not all upvotes and downvotes are weighted equally. Upvotes from people who have more upvotes count more. There are additional details to prevent some of the more obvious ways to game that system.
- ssivark 7y ago> Upvotes from people who have more upvotes count more. While this will avoid the problem of "junk opinion democracy" (each voice gets one vote, whatever the expertise), this still wouldn't avoid the problems attributed to the scientific establishment, where authority is roughly proportional to impact/prolificness/citations.
- lisper 7y agoThe difference being that in what I have in mind, anyone can publish and anyone can review, so it would be more like Arxiv and less like Nature or Science. In the scientific world there are stringent filters in place before you are allowed to play the game at all. Also, scientific communities tend to be small, and they are dependent on each other for funding. That introduces politics and perverse incentives. (I used to be a researcher. The politics and incestuousness is one of the reasons I quit.)
- abecedarius 7y agoAre you familiar with https://en.wikipedia.org/wiki/Advogato https://en.wikipedia.org/wiki/Advogato and similar work from the early 2000s? I still think this kind of thing has potential.