I think the upvote/downvote model for evaluating content is fundamentally not a good one, and this deficiency is the root problem for a lot of sites (Reddit, Quora, etc.) that were founded around the same time, circa 2005-2012. It lends itself too much to groupthink, popularity contests, and self-promotion.
X/Twitter has a million other problems, but I do like their new Community Notes of fact-checking claims. It's essentially anonymous and somewhat immune to self-promotion.
Going forward, a better path for FAQ-type sites is probably something similar to that model. The hard part is how you get users to answer questions while removing the self-promotion and gamification/scoreboard incentives.
It's possible that AI might just destroy this space entirely, but I still think you need some form of human fact-checking in order to avoid hallucinations.
>I think the upvote/downvote model for evaluating content is fundamentally not a good one
This is why the bump model that 4chan uses is superior. You can sort threads by activity, but all posters display with equal weight, and you must utilize critical thinking to evaluate each and every post (unless it is from a trip user whose authority you trust in a particular domain), not some little number that pushes them to the top or sinks them to the bottom. There are (You)'s, but that can indicate consensus, disagreement, controversy, or something else, and you need to sus that our for yourself.
Consensus-based truth models are frankly shit for most types of topics. Heck, even for scientific topics, the peer review model has been shown to be completely bankrupt.
> This is why the bump model that 4chan uses is superior.
The elephant in the room is that 4chan's average quality of content is by far the worst of any major forum platform. The "bump" model is also used iirc by most bbs or older-style internet forums. In the modern age, incel forums are probably the most popular of these older-style forums (and surprisingly they probably have only slightly better average quality than 4chan).
So, correct me if I am wrong, but the bump model usually results in shitty average quality of information.
As the bump model needs heavy moderation as the upvote/downvote system is a form of self-moderation/censorship. However, the entire appeal of most sites that use the bump model is to shitpost or be edgy...
Honestly, the bump model does sound intriguing from a utopian, critical-thinking first, free speech first-principles perspective, but I'd like to see an example of a successful popular forum that implements such in reality.
The bump model allows chaos wherein you, the reader, must critically evaluate everything with no preconceived notions or pre-filtering. This is how you question, re-evaluate, and sharpen your knowledge through productive debates. There are diamonds and gems in "flat", bump-structured fora, that do not exist in other group-think, mass-consensus-based sites.
Why does the average quality of information matter? You are making a trade-off for convenience. It is indeed convenient to accept common group-think. If you are critical enough to recognize low quality posts on bump model sites, why are you concerned with low quality posts? Read on to the next post.
Insert the meme here of Cypher stating that he doesn't even see the code anymore, but the remixed version of "I don't even see the shitposts and shills anymore."
>I'd like to see an example of a successful popular forum that implements such in reality.
You know what they are, but I will elect to avoid an instaban by posting them here.
> You know what they are, but I will elect to avoid an instaban by posting them here.
You cannot list them, because I already have and I have not been instabanned.
There is not actually much chaos in these supposed "bastions of free speech". Because in practice, threads are often created by highly motivated users and can be brigaded by coordinated groups (IRC/Discord groups). Even without so, culture and groupthink emerges simply because people mimic each other.
This is the problem with all utopian first-principles liberty thinking, in practice, they fall prey to the same easily observable defects in human nature. Thus, forums like hackernews thrive because of competent moderation in their stead.
High average information quality is a top 3 important metric for browsing any content on the internet, and quality does not "emerge" effectively through this supposed "diamond in the rough" system.
It comes down to the same thing at the end of the day. You gotta pay for quality (be it quality contributors, or simply quality fact checkers to properly moderate information) and no one wants to pay. Not the users, not the webmasters. So the best way is to leave it to the users to sort out. We know that is error prone, but at least works some of the time.
Community Notes could just be an expansion of a good tweet to clear up some ambiguity.
The up/down feature tends to create echo chambers that eventually exclude half the population. Reddit is worse at this because of excessive moderation combined with doomscrollong incentives.
X/Twitter has a million other problems, but I do like their new Community Notes of fact-checking claims. It's essentially anonymous and somewhat immune to self-promotion.
Going forward, a better path for FAQ-type sites is probably something similar to that model. The hard part is how you get users to answer questions while removing the self-promotion and gamification/scoreboard incentives.
It's possible that AI might just destroy this space entirely, but I still think you need some form of human fact-checking in order to avoid hallucinations.