How Wikipedia became the last good place on the internet

The next best thought out knowledge-based site is StackExchange.

I think it's a good rule of thumb to not have people edit articles too close to them, but for some objective facts you may as well let it in (and remember: "facts" means "can be proven OR disproven". I can make a fake birth certifricate but you can objecively prove through later talks that I lied. Or the hospital realizes they messed up and my own birthday is a lie to begin with. Much easier to reason with than "am I a good person?")

[Accessing Wikipedia from command line I never see the nags for donations.]

Advertising provides a perverse incentive that has the effect of injecting rot into the core of products that depend on it.

Some jumped-up self-appointed 'editor' appeared as-if-by-magic, made an obviously stupid decision to keep all the copy pasta, backed it up with a kangaroo court of 1 or 2 buddies weighing in on the Talk page, and that was that - 30 polluted pages.

Many times, what you actually experience is a variant on the smart mob concept [1], which are enabled by watchlists and diverse central coordination pages[2] . The number of ways you could organize people on wikipedia without them even realizing (let alone being offended by it) by smart use of smart-mobs was (and probably still is) endless.

Of course, for a casual observer, these smart mobs look very much like coordinated cabals, even if in reality the individual members don't even know each other.

I'm not sure how well this base principle has been documented though. Possibly

[1] https://en.wikipedia.org/wiki/Smart_mob

[2] eg DRN, ANI, AFD, diverse wikiprojects:

   * https://en.wikipedia.org/wiki/Wikipedia:Dispute_resolution_noticeboard

   * https://en.wikipedia.org/wiki/Wikipedia:Administrators%27_noticeboard/Incidents

   * https://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion

   * https://en.wikipedia.org/wiki/Wikipedia:WikiProject

So just throw all theory/hypothesis in the same bucket because one might be true? You can't cherry pick one thing that happened to be true and say "it is wrong to label that pseudoscience" There are a lot of bad theories and hypothesis out there.

Clathrate gun hypothesis is just that, and should be labeled as such.

I read wikipedia articles about phenomena my academic discipline is involved with and I keep thinking, "why do they still push that as true? This is ridiculous" about every time I log in.

But it serves hegemonic knowledge production,so it's in there, and we now know the effort the US government is making to make sure its viewpoint gets shown as "true". It's disgusting at times.

The plate tectonics hypothesis was criticized because there was no known mechanism for how continental plates could drift. Until that mechanism was understood, it was perfectly good science to be critical of Wegener's hypothesis, and a Wikipedia of that time would have been entirely correct to document the hypothesis as controversial. Hindsight is 20:20.

It's also very much not an encyclopedia's job to try to discover the "truth", whatever that means! Its job is explicitly to document current consensus understanding, however flawed that might be. Trying to figure out the "truth" is very much above Wikipedia editors' pay grade.

Pseudoscience and conspiracy theories are something entirely different from simply proposing hypotheses that might not be backed by currently available evidence, and it's incredibly fallacious and dangerous to try to equate homeopathy or flat-Earthism or climate change denial with something like hand hygiene or plate tectonics.

One thing I see with 100% of conspiracy theorists is that they have no self-awareness of when they're wrong. They will believe 1000 things and when 1 of them turns out to be true, they both think it validates the other 999 and will never consider they were perhaps right for the wrong reasons. My point is that, as you say, "Being firmly anti-pseudoscience and anti-conspiracy theory and anti-fringe might not be the best heuristic for discovering the truth", but it's the best we have. I'll also paraphrase one more quote that I heard from the great Tim Minchin, but he didn't know who to credit it to, "if you open your mind too much, your brain will fall out".

https://en.wikipedia.org/wiki/Modified_Newtonian_dynamics

If anything, my criticism of this article is that the first few paragraphs don't explain how fringe it is. (And I'm firmly on the opinion that working on it is important, because it's the only path out of the almost disproved box that we have.)

If Wikipedia existed in its current form in the mid-20th century, it would have dismissed plate tectonics. But, importantly, as the scientific consensus evolved, so would Wikipedia's stance on the theory.

This is the key difference between science/academia and politics/religion. The former changes with new evidence while the latter doubles down in the face of it.

That's true, but it's fine if Wikipedia, not being able to discern truth and falsehood as "a collection of people arguing online about stuff," uses consensus as a low-resolution proxy. It's the best tool they have given their circumstances.

> Today's conspiracy theory could be tomorrow's government press release

But it usually isn't, which is why consensus works as a pretty good low-resolution proxy for truth.

I am almost certain they would of. The "fringe" and "pseudoscience" labels are used much more broadly on Wikipedia then one might expect.

I have seen a Palestinian scholar, whose research went against the prevailing Euro-centric Western view, labelled as "fringe" on Wikipedia even though it was being published by UNESCO and Cambridge University Press and he was by every metric a mainstream academic. The main editor's critique of his work was mainly ad-hominem and that it hadn't been acknowledged or "approved" by Western scholars. He didn't seem interested in verifying the texts that he cited and quoted; texts that the previous scholars were ignorant of. This was simply a case of a likely correct but minority view that was also difficult for an English speaking Westerner to verify, as the cited works were obscure and written in the 9th/10th century in Arabic. One of the few known manuscripts was located in a religious building in Turkey and not known to be available online in either scanned or transcribed form.

For the longest time, Wikipedia simply presented facts, with sources, and allowed the user to make their own judgments. Nowadays, many pages are so heavily and obviously biased in one direction.

A personal example I noticed is when I did research on Intelligent Design. The page for it is so heavily biased in one direction, the idea that it’s completely BS “pseudoscience.” That may be the case, but how about you let me read the article and come to that conclusion myself, rather than poison the well from the start, by attempting to bias me in one direction before I’ve even read anything beyond the summary.

Jerks found other ways to be jerks?

FWIW, Wikipedia also has this problem. Just spend half an hour going through some page histories and browsing some user pages, you'll find feuds and infighting all over the place. People with a profile saying "I stopped contributing because user XYZ essentially bullied me", yadda yadda.

Well, that failed spectacularly. The Stack Exchange sites are some of the most hostile, caustic places on the Internet, surpassed on that axis by only a small handful I'm aware of.

Also with HN, many subreddits, millions of personal blogs and sites...

I tried to find why the paper says that Wikipedia is "the last good place on the internet", but it doesn't seem to provide any reference for that. Is it a meme somewhere?

Every time I have posted a question on SE it is either flagged as already answered (when it's not) or people screaming "XY Problem!"

My approach to SE now is "Look, Don't Touch".

I love this idea. Not likely to happen with Wikipedia as it currently exists, but could possibly be built with a Wikipedia article as one of many links listed for a particular subject.

Slight tangent, but this is why I go to All Sides News when I feel I need more varied perspective on an issue. I don't use it a lot, but when I do I'm super glad it exists.

Came here to say this. 98% of the content is perfectly fine. Searching for "lemurs" in any search engine returns a perfectly trustworthy (calling all lemur experts!) wikipedia page. But, look for anything divisive, political, cultural, etc., and I've increasingly noticed the articles take clear positions.

I barely recognize much of the 70's and 80's history on wiki. A lot of it clearly wasn't written by someone who lived through it, and if it isn't linkable it basically doesn't exist. History is written by the victors, and they only do the internet. All those books, newspaper and magazine articles of that era just don't exist outside of libraries and microfiche and it's clear none of those wiki editors are going there.

I've seen younger people start to complain about this as link rot reaches into their generation.

At least academics have standards, even if they don't always follow them.

I wrote countless articles for them, but they didn’t give a crap. I now contribute photos to commons.

Don’t bother contributing to Wikipedia. Only people who fiddle around with changing URLs, categories and templates are appreciated. It’s not worthwhile putting in the effort.

It sounds like if you joined any open source, open collaboration project, or volunteer organization, you'd be a huge pain to work with unless everyone agreed with you all the time.

https://gutenberg.net

https://www.pgdp.net

it's one click away with uBlockOrigin. disable for all and enable per domain as they deserve it.

hit a link that don't load? hit back and try another. it's not like there's no alternatives for everything.

(I know the name is a bit of a retrofit from after O'Reilly decided to host a web 2.0 conference. Back when it was meaningless.)

(I have zero idea about the actual numbers)

One reason that corporate internet won is that money transfers to a single entity, and that pot of money can be used to coordinate action in some particular direction (for better or worse depending on your view).

Without some coordination mechanism, it seems a a bit too divided. It's unclear to me what the appropriate mechanism is, but I think that's the right sort of way to frame the problem.

What we need is a social-first internet where every piece of content is associated with an actual human and people query for content through their social graph. This will be doubly important in the age of AI.

The Epsilon galaxy is a friendly outpost of services in the brave new decentralized digital universe, now known as the Fediverse. (I coined the term Fediverse “galaxy” to refer to a cluster of federated services/nodes run by a single entity, and it’s gained some traction on Mastodon).

My galaxy is completely bootstrapped, no VC investment: https://epsilon.lol

We even have our first paying customer (8 euro/ year)!

If the internet was getting way worse in practice wouldn't a bunch of people start using somethng else and new user rates decreasing? By all metrics the internet is the best it ever was and keeps being super useful for everyone.

Yet when I open the explore page on Mastodon I still see the same shit: politics. https://mastodon.social/explore

It's almost as if the whole point of Mastodon wasn't anything more than just to be another echo chamber. But hey no bad guys from "the other side" so guess we good.

HN is still the best. We need more HN not Twitter or Mastodon.

The license provides a wide umbrella so that many different people and entities can use the content without worry. Contributions provide a fly-wheel of improvement.

A lot of the current batch of internet companies provide walled gardens and eventually succumb to the trappings of monopolies. In some sense, I think it's survivorship bias. The only entities that stand the test of time are the ones that provide content under libre/free licensing.

While I don't disagree with your list, I think a missing component is access to the commons. Without a rich commons, our society suffers as content sinks into a copyright quagmire.

Part of the early internet was about being able to find a huge range of different attitudes and beliefs, every possible politics, every possible religion. It was about being able to get a glimpse of non-Western cultures you might never have been able to learn about before. There is very little diversity in the fediverse.

Lemmy and Mastodon are good examples of this. Federation is a PITA. I know people who tried to switch from Twitter to Mastodon and never succeeded, getting lost in the concept of different instances etc etc. Mastodon instances are relatively expensive to run because the day to day operation of federation is way more complex than a centralised solution would be.

There’s nothing wrong with centralised services and Wikipedia is a great example of one that works. You just need the right governance structure and an audience willing to chip in to cover costs. Federation is over complicated and often unnecessary.

Humanity's biggest issues are coming from concentration or centralisation. You can apply this pretty much to anything: wealth, population, corporations, etc. We create bubbles based on these.

The solution is decentralisation.

The question is how to get there?

* I don't use any social media except here. * I have nothing to say so don't host * Duck duck go? * I tell everyone to move to firefox, help them set it up and most are convinced.

Just wish I knew a way to actually help.

There are good reasons the fediverse hasn't caught on yet and it comes down to user experience. We haven't figured out how to build a federated service that truly coheres as a single platform, network, or community across the whole network. (Yes, individual instances act as coherent communities, but the links to other instances are weak.) And we haven't figured out how to reduce the cognitive load of joining a federated platform to the point where it's accessible to the average user . I think this is a big part of why Mastodon is still touting usage in the single digit millions while Threads and Bluesky grew several orders of magnitude more as Twitter melted down.

Until we solve those problems - which by necessity includes solving the discovery problem - we're not going to be able to successfully decentralize the net. Those problems are a huge part of what lead the net to centralize into massive platforms in the first place, and we can't undo it with out resolving those factors.

And solving coherence and discovery in federation is a multi-trillion dollar, decades and a research team, Very Hard Problem (tm).

...but we know how to build centralized platforms that are democratically governed by non-profits. Wikipedia showed us how, and there's a lot that could be improved upon to do even better in terms of transparency and governance. And yes, democracy is hard. It just is. Community is hard. Society is hard. It requires deliberation and compromise. And it will never be perfect (as Wikipedia is far from perfect).

But it would sure as hell be better than what we've got.

The main piece that's missing is startup funding. Non-profit funders mostly don't understand software. We need non-profit funders who are willing to fund the development of platforms until they gain the sorts of critical mass that would allow them to fund their operations through donations.

When we pay for services, we can fight spam/ad invested sites, at least to some extent

You know what services we have that are federated? Phone (POTS), SMS and email. You know what services have a massive problem with spam, fraud, false positives on detection of the aforementioned, ID spoofing, etc? The exact same services.

People ignore the real reason Wikipedia is successful: ownership. It's also worth noting that Wikipedia isn't federated either.

We have the same issue with ISPs in the US. The best Internet is municipal broadband. The national ISPs are pretty much universally despised. Again, what's the difference? Ownership. With municipal broadband, the residents own the broadband. National ISPs exist to extract wealth from customers while doing the least possible and lobbying for legal protections against competition.

It's almost like the best results are achieved when the workers own the means of production. Someone should really look into that.

What the web needs is new layers to make what your suggestions much more practical.

Creating a website was a nice idea back in 2003 when mostly geeks with skills were using the web. Today, most people don't want to put forth the effort of creating a website. Many geeks with skills that are still around today don't want to bother creating a website because few people will find it. If you make a website, at best only people who read websites will find it. A growing number of people today don't even know what a "website" is because they hardly live on the web proper. Free web hosting is generally bad news because the platform is far more likely to either shut down or provide zero support.

So back to the thing I said about new layers to the web...

If we're going to bring back websites, then we need to bring back the idea that sites will be hosted on many individual machines that aren't necessarily server infrastructure. With the bandwidth most ISPs and devices are capable of today, data isn't a limiting factor in terms of serving a standard website to even thousands of users. The web was meant to be closer to a peer-to-peer model, and it can be used peer-to-peer, but the layer just isn't there to make that model seamless. Most P2P protocols avoid fiddling with the actual network stack and do entirely their own thing; BitTorrent for the most part lives in their own client applications, as historically have many other P2P protocols. That was fine when geeks were using them, but this is insufficient for making the web itself P2P.

The closest implementation to what I am describing here is Yggdrasil:

https://yggdrasil-network.github.io/

Yggdrasil implements P2P mesh networks by behaving like a VPN, assigning an IPv6 address to peers based on public keys. There's no one network, so people are free to create their own networks, and the same address can be used between networks. Although it has some similarities with anonymity networks like Tor, it is not designed for anonymity, which I think is a good thing because Yggdrasil should focus specifically on the networking aspect. Although it's technically in alpha status, I've already been using it to network devices together in my home office. I even have it installed in my car's Android head unit so I can SSH into it and do Linuxy things with it.

The current issue I see with Yggdrasil is that it's not plug-n-play enough to become the de facto next generation layer of the web/internet. It still requires manually configuring a JSON file in order to join a network and so forth. Yggdrasil isn't difficult to install, but if the world wants to go with something like it as this next layer, we need to get it to a state where the user needs next to no knowledge to set it up.

With this hypothetical layer, groups of people can form their own networks and exchange addresses. This is perfect if you want to have an application similar to WhatsApp or Signal that is mostly private and scoped to groups of people, but allows users to cross into multiple groups.

In terms of websites, it's perfect because you get an IPv6 address that is publicly available to anyone on your network and with none of the hassle that's involved with getting a public IP address on the clearnet. It's all a matter of who is connected to the network.

If you want anonymity, this layer can still work through another layer like Tor, or even live entirely within something like I2P. A benefit to using I2P here is that its design makes DDoS attacks more difficult, at least in theory.

With such a layer, hosting sites can become much more trivial. For everyday people who are a little more technically savvy but don't want to be fussing with the command line, I picture simply downloading something like a Reddit clone that runs like any other app on Linux or MacOS and it runs with a neat little admin interface. Share the address (or the domain if a secondary DNS is used) and you're off to the races.

You have the power mods issue, only far worse, because they are vastly more powerful now that they control whole instances.

Then you have all the fascists who claim they are leftists but are actually just "anything Anti-USA is amazing". They also run major instances.

I ended up deleting my instance, returning to Reddit (and simply using it from desktop only, and regularly delete my content), and the quality of discourse is already so much better again. I did create an account on beehaw and treat it as unfederated (only subscribing to local communities), the people there can be a bit strange, and there’s not a huge amount of content, but at least they aren’t as bad as the rest of the Lemmy fediverse.

Like before Eternal September, people had to learn stuff to do things on the internet.

Also in response to your edit, isn't the entire point of Hacker News about discovery? One thing we could do right here on is have a weekly thread for people to post general places on the internet they find interesting (webpages, services, communities) but perhaps don't feel like they constitute a unique thread of their own.

Suppose your webbable hobby is curating trustworthy product reviews. You're the hero that's going to fix the web AND capitalism. You can still do that, but you're going to help fewer people because you have to split your attention between you know, doing the thing, and doing the kind of marketing which helps people tell the difference between your thing and the ones that are secretly ads. Meanwhile, the bad guys can focus entirely on the latter. Sorry about the time you wanted to spend comparing blenders, your hobby is now competing with marketers... Or you have to make peace with either obscurity, or contributing to the thing you're trying to kill, because you know that the second you start making an impact somebody is going to rehost your site interleaved with misinformation.

Yes, let's fix this thing but the answer is protocols with different inherent priorities. We can't shout over the noise, we need to build something less noisy.

My personal view is that we should stop building trust around server names and implicitly trusting whatever those servers send us. In this model the non-content-generating work converges on systems to identify and remove malicious content (either from the server or from the browser screen).

Instead we should be addressing content, and sourcing it from whatever device has a copy and is most near. Our browser should know which people we trust and should make it easy for us to lean on that trust such that we can propagate only the trusted bits in the first place.

That is: stop identifying ads so that we can block them, start identifying trustworthy content so we can all participate in propagating it more efficiently and supporting it's creators.

Or to put it differently: there needs to be no way to spend money to promote content. No third parties injecting bits anywhere. If you notice that kind of shit, you apply annotations which distinguish the signal from the noise and then you propagate only the signal. If people abuse this power, you revoke trust in them and move on.

It'll require an awareness of information hygiene, which will be work to maintain, but all of the alternatives I can see involve tolerating life in a Skinner box

Even so, I've managed to edit the Derek Chauvin article without any trouble. There's a certain shape of controversy that's painful to get involved with.

Usually it's matters of nationalism. For instance, anything with respect to India is a complicated thing to act on.

This is exactly what I do. When I visit articles about politically discussed topics I often read the Talk pages. I usually find important and (regularly) well-sourced information on there that, for some reason, do not make it into the article.

Lack of viewpoints is the least of the problems with "traditional media". Fox news and MSNBC have the same, debilitating problems as NBC, CBS and ABC.

1. They cover things poorly, little depth 2. They are beholden to their sponsorship (a great non-political example is their unwavering support of particular CFB conferences) 3. They offer opinions first instead of facts 4. They are highly selective of only certain topics (Wikipedia is really broad despite having moderation) 5. Wikipedia has 120k active editors. Networks are lucky to break 1000 employees.

I think that US traditional media could do a better job, they've just catered to the LCD. There are examples of non-US media that do a solid job of maintaining NPOV and cover topics in reasonable detail without injecting overt opinions.

Your local library probably has a magazine rack with some stupid shit in it but is it not stupid shit that patrons are interested in? Wikipedia ousted it for better or for worse.

Even if those things end up being correct, it's also not a place for original research because nobody who edits Wikipedia is qualified to make the kind of judgment on whether original research meets any meaningful quality standard. Submit to a journal if you think you have a worthwhile revolutionary idea. It'll make it to Wikipedia after it becomes generally accepted and that's fine. I don't think a single knowledge repository can realistically hope to be both a repository of trusted, accepted knowledge of the day and a place to hash out the ongoing debate about what should be accepted knowledge and where we might be wrong.

This may be true in the US, I wouldn't know, but I work in media, and that hasn't been my experience at all.

They do as good of a job as any of us in trying to stay neutral.

To elaborate, I think that this obsession creates a few interesting scenarios. If people are flatly told/believe how a given media is biased then they don't have to read critically themselves. It can also be used to shut down discussion, your source is biased- I can ignore it.

> Yeah, everything is relative, truth doesn't exist, yadda-yadda.

strikes me as downright ironic. The routine and wholesale dismissal of sources because they don't tell the story with someone's preferred angle is amounting to a whole lot of yadda-yadda, I agree!

Well, Google didn't exist until 1998, and hadn't made much impact until about 2002.

For politicized stuff I see it no differently than any other aggregator like Reddit or Hacker News. get a quick feel from the post and comments (yeah I even read the editor comments), look at several sources and get a feel for their bias. Then try to form my own view based on that. I see bias/editorialization there because I see it in every single person's POV. That's just life and why the number of sources I'd ever take wholesale are ones I can count on onehand.

Wikipedia is definitely very weak at developing stories, where you'll really see the back and forth happen underneath.