[thanosnose]

> Something needs to change, the same way Reddit’s ranking algorithm was changed based on Randall Munroe’s public suggestion.

That was a long time ago. Reddit ranking is even worse than HN's now and is manipulated by political and news organizations now with the help of reddit employees. Reddit is a highly biased propaganda site. It's not a platform anyone should be using. In many ways, reddit is worse than facebook.

I stopped using reddit once I could predict what I was going to see on the frontpage day after day. Sadly, I'm starting to be able to predict what's going to be on HN's frontpage now.

[EGreg]

It’s a very serious area of research. Are there any good solutions in the literature to collusion-resistant voting and reputation?

Something like hotornot is voting resistant because it shows RANDOM photos so it would take a lot of effort and time for people to collude to upvote a particular photo.

But that is a site about one topic: hotness. How do you get people in smaller forums rate eg the strength of an argument without causing people opining on stuff they have no idea about (eg career politicians or chess fans rating claims about molecular biology).

I am seriously interested in building a site where each claim is debated exactly ONCE - you come and either upvote a pro/con argument (which itself rests on other claims) or submit a new one, and it would use Randall Munroe’s ranking system for display.

The idea is to make people discuss the truth or falsity of each claim once and for all. Politics, religion, news etc.

I can see how evidence ca accrue, eg if a video of an attack in Yemen is uploaded, it would join a growing list of other videos claiming to be accurate representations of this event. Each video should be kept. But which woukd be upvoted? What if some are doctored and heavily upvoted in a coordinated attempt by “sleeper” accounts that behaved well until that point? How would HN detect it?

Believable fake videos are not (yet) able to be produced. But fake photos already are. And other claims can be pure garbage. Fake news of all kinds has permeated the Web. I have often said that our systems have been designed to rely on the INEFFICIENCY of an attacker, and that needs to be changed in the next systems we design.

Voting is one of those. I understand how to make voting secure private and verifiable with Merkle Trees. I don’t know how to prevent collusion.

And any site which purports to bring everyone together to determine “truth” becomes a honeypot for disproportionate efforts to game it.

How does Wikipedia stay relatively good? At least as good as Britannica? I would love to more collaborative (open source) instead of competitive (capitalist) systems for software, news, drugs as it is for science etc.

How does science do it? They don’t, sometimes.

When it comes to centralized beliefs about truth, whether it’s who owns a token (preventing double-spending) or whether Gandhi “really” said that quote or whether Tienemen square really did happen, what is some literature about consensus about REAL WORLD (not electronically verifiable) claims that actually can’t be gamed so easily?