Live data from Hacker News

Ask HN: How can we fix Wikipedia?

news.ycombinator.com

111–120 of 301 posts

Re: Ask HN: How can we fix Wikipedia?

#111

I'm not sure if I share the same assessment. I am amazed at how robust your typical Wikipedia page is despite the chaos that goes into writing it. Given that it's not possible to make a version of reality that everyone is happy with, I think Wikipedia comes pretty dang close.

This is a good example of an amazingly bad article with the worst type of wiki lawyering on the talk page. It basically still has the same conclusion in the lead as when it was originally created, and most of the claims/leaks against Hunter were considered "false"/dinsinfo... But now that those laptop leaks were either confirmed by different reputable sources, it more or less indirectly calls most of the claims true…

> the extreme selectivity in sources somehow just... disappear

Unimportant articles will draw a different crowd than more important articles. Those different groups will have different standards for sources and tone/bias.

What changes have you made to that article that were rejected / rolled back? How long have you been making edits to it?

Re: Ask HN: How can we fix Wikipedia?

#112

Get the Wikimedia Foundation to spend money on the technical foundation of the sites, rather than spraying millions up the wall of "racial justice". That's not me making a strawman by going straight to "omg, the SJWs", it's a exact phrase from the 3rd paragraph of the 4.5 million Knowledge Equity Fund page [1]. That, and executive pay, is where donations go, not to keeping the lights on. The running cost to keeping t…

This could almost be taken word for word from an article about Mozilla. People who think 'social justice' is a bad thing seem to set really high standards for the open, independant organisations they say they support, and they're not afraid to give them some extreme tough love and undermine them at every step if they don't follow their priorities.

They didn’t get my donations saying they’re gonna bring about racial justice. They said this is what is needed to keep Wikipedia without ads and that they’re suffering. Misleading is an understatement.

Re: Ask HN: How can we fix Wikipedia?

#113

Get the Wikimedia Foundation to spend money on the technical foundation of the sites, rather than spraying millions up the wall of "racial justice". That's not me making a strawman by going straight to "omg, the SJWs", it's a exact phrase from the 3rd paragraph of the 4.5 million Knowledge Equity Fund page [1]. That, and executive pay, is where donations go, not to keeping the lights on. The running cost to keeping t…

I was ashamed when I learnt Wikipedia artificially funds the writing of pages of women, so they try to reach equal numbers to men. I wouldn’t be ashamed if the criteria for admissibility were equal, but barely being a researcher with published papers is enough, while men with entire books on the subject get rejected. Talk about rewriting history…

> while men with entire books on the subject get rejected.

Ok, I'll bite. Giving incentives to increase the number of pages about women, does not impact in any way the number of existing pages about men.

Re: Ask HN: How can we fix Wikipedia?

#114
So Wikipedia has a bunch of problems, it looks like you are focused on the political problem. By this, we must be careful to say that we don't mean the American political problem that is embedded on Wikipedia[1]. The problem you are having is that a particular Wikipedia political system exists separately from all that. Like, there is an artificial politics of contribution, there are committees and meetings and weird agreements about this is acceptable but that is not, elections and moderators and so forth. That is the problem you identify. Call it a LARP. You wanted to contribute or correct information and you find yourself LARPing as a senator to get it through.

Open source software gives a way to address those political issues. So, they exist in OSS, some open source projects get these elaborate governance structures, they invent a governance LARP with elections and democracy etc. But some OSS projects have not had this.

Cantrill offers an interesting insight about this in his talks... He points out that merge algorithms give a powerful way to fight this, because once you have a merge algorithm you can allow two realities to exist simultaneously. He says the LARP is a result of being “forkaphobic”, afraid of allowing the project to splinter into two identical projects. Whereas, he says, if you look at the Linux kernel, Torvalds can tell you you're full of shit and you can do the thing anyway on your own branch and people can be like “that's actually really interesting” and Torvalds can admit he was wrong.

So the fundamental thing you have to solve is allowing an article to exist in two independent states that get merged together at the end. And if you can solve this then you do not need a political structure to get everybody to agree because your technological process tolerates disagreement.

1. By the American political problem I mean this particular construction of right-vs-left where divisiveness reigns and nobody agrees on even basic facts... there is some of that, but it's not the problem you're having. 2. Some other problems include:

- Wikipedia articles are typically extremely long. This is necessary because they do not have hierarchy but instead kind of cross-link to their peers which prevents any other way of attaining good depths in education. Information is routinely repeated in many pages.

- Encyclopedias are biased to a sort of declarative information, whereas in most cases you really want to provide imperative information: think of what would happen to a university if we forbade labs, recitation sections, homework/exercises, practice tests, instead you must learn everything from lecture without ever assembling a circuit yourself. Who is gonna learn electrical engineering there? But Wikipedia is a fount of information about electrical engineering components, mathematical topics, there are physics articles. Sigh.

- And then in terms of accessibility, Wikipedia has not solved the hard problem of allowing users to distribute computations in some vulgar programming language (Excel, HyperCard) similarly to how they enable typesetting in a vulgar markup language. As a result there is nothing interactive, there is barely any video or audio, everything must be textbook-serializable. Which makes it hard for us to have a volunteer effort of, here is a 150-word card on a topic, here are the deeper dives you can go into, here is how you can interactively play with this concept, here is some volunteer reading it to you so that you don't need a screen reader, here are examples to test whether you have understood the idea, etc. You want the same idea to have many different perspectives.

Re: Ask HN: How can we fix Wikipedia?

#115

I'm not sure if I share the same assessment. I am amazed at how robust your typical Wikipedia page is despite the chaos that goes into writing it. Given that it's not possible to make a version of reality that everyone is happy with, I think Wikipedia comes pretty dang close.

This is a good example of an amazingly bad article with the worst type of wiki lawyering on the talk page. It basically still has the same conclusion in the lead as when it was originally created, and most of the claims/leaks against Hunter were considered "false"/dinsinfo... But now that those laptop leaks were either confirmed by different reputable sources, it more or less indirectly calls most of the claims true…

That makes me wonder what is the use in including anything political except blunt facts for a place like Wikipedia. If they want to do “political history” then I’d just have an open dump site of any kindling people want to add to the “discussion” and leave it at that. Basically just an archive of propaganda at worst.

Re: Ask HN: How can we fix Wikipedia?

#116

Get the Wikimedia Foundation to spend money on the technical foundation of the sites, rather than spraying millions up the wall of "racial justice". That's not me making a strawman by going straight to "omg, the SJWs", it's a exact phrase from the 3rd paragraph of the 4.5 million Knowledge Equity Fund page [1]. That, and executive pay, is where donations go, not to keeping the lights on. The running cost to keeping t…

From the first link:

>The Wikimedia Foundation defines racial equity as shifting away from Eurocentricity, White-male-imperialist-patriarchal supremacy, superiority, power and privilege to create an environment that is inclusive and reflects the experiences of communities of color worldwide.

What an interesting signaling statement. If you did not know what "racial equity" was, imagine trying to parse out what it is from that statement.

Imagine reading this phrase out loud to anybody not already involved in this debate. Would this help them understand what "racial equity" was? If you read this statement to any of my family or friends, who live in the middle of the country, they would probably conclude that this was some sort of racist propaganda thing, and wouldn't want to talk to whoever brought it up anymore.

I mean the fact that the foundation which is supposed to also be providing information to the world is also proudly publishing things like that should be pretty revelatory.

Re: Ask HN: How can we fix Wikipedia?

#117
post #31
post #11

Sorry, I am not quite sure what the problem is? I think Wikipedia is great from my day to usage. Can you point me to specific examples where the output is bad due to this? I would like to see the problem.

I personally haven't seen a biased Wikipedia article in a very long time. It is absolutely astounding to me and seems like one of the great successes of internet culture. I assume OP has a particular set of issues that they are not in agreement with the majority of experts on.

>I personally haven't seen a biased Wikipedia article in a very long time.

What leads you to believe you can so easily spot a biased article? Everyone believes they're a neutral arbiter.

Re: Ask HN: How can we fix Wikipedia?

#118

I think overall the wikipedia model has shown itself to be decently working. But (anecdata) I have my gripes with the way it handles evidence. My SO has a person (died already) in her family with a (relatively) short wikipedia entry. There are factual errors in the article. We know that, because we have original documents showing that the "facts" reportet in the entry are wrong. Due to the fact that the documents can…

Now more than ever, Wikipedia must be publicly verifiable. That means that any claim can be checked by anyone with an Internet connection and a library card. Imagine if contested articles contained controversial facts on the basis of secret evidence? The editor made the right call.

But if the editors receive proof that a statement article is wrong, but that information can't be published on the Internet for copyright or other reasons, that means it is proven that the source of the information is wrong, so that source should be removed as unreliable at least in this case. If the rules of Wikipedia don't allow unsourced information, they don't prevent removing incorrect information that has been shown to be incorrect. So they could nuke the bad information. If, after that, nothing is left, they can remove the entire article.

Re: Ask HN: How can we fix Wikipedia?

#119
post #104

You can't. Anything that is editorialized will skew to an extreme over time. In the past, wikipedia was bad because its quality of information was generally terrible. "Alan Turing proved in 1936 that a general algorithm to solve the halting problem for all possible program-input pairs cannot exist." https://en.wikipedia.org/wiki/Halting_problem Alan Turing did no such thing, but that's wikipedia for you. On topics yo…

> Anything that is editorialized will skew to an extreme over time

It is exactly why Actuator’s point in the OP was actually incorrect.

> While Wikipedia's model seems far better than an individual publishing house's encyclopedia

The illusion of an unbiased “encyclopedia for the world” has harmed the decentralization of information. It has actually created a unipolarity much like we see elsewhere around the world.

It would be far better to have multiple different encyclopedia companies that could compete and perhaps even specialize in certain areas.

But how do we keep the convenience of having them all connected? That is what Larry Sanger has been working on[1]. A federated network of encyclopedias that use the same data formats.

[1]: https://encyclosphere.org/

Re: Ask HN: How can we fix Wikipedia?

#120

Earlier quoted context omitted.

> First is that "Reality has a well-known liberal bias". Uh...what? Where is the proof of this?

He is quoting political comedian Stephen Colbert, referring to the fact that different political groups all independently believe that their perspective is actually a reflection of reality.

To the contrary-- his neocon character is admitting that his perspective does not align with reality. That's a much more scathing critique of the neocons than the generic upshot you wrote above.
Post reply on HN