Live data from Hacker News

25 Years of Wikipedia

wikipedia25.org

191–200 of 498 posts

Re: 25 Years of Wikipedia

#191

I do believe that Wikipedia is one of the least biased sources out there, but there is definitely bias. Here is a concrete example. Compare the introduction paragraph of the English circumcision article ( https://en.wikipedia.org/wiki/Circumcision ) with the German one ( https://de.wikipedia.org/wiki/Zirkumzision ). The English intro talks a lot about medical advantages of the procedure: "reduced rates of sexually tr…

It also shows that it isn't perfectly organized, that it isn't an ideal model for knowledge aggregation. If it's ideal for it to be globally consistent, then it doesn't have that. If it's ideal for it to be adapted to different cultures, then it doesn't have that either, because the divisions are based only on language. However, Wikipedia it is really an amazing place, and it should continue to be preserved and improved.

Re: 25 Years of Wikipedia

#192

Earlier quoted context omitted.

Oh dear, you need to learn about the GamerGate incident which started August 2012. All the extreme division and online manipulation through the collaborative creation of false narratives started right there, with that issue, before contaminating the entire political landscape. It's the Eternal September of our generation, and it's not recognised enough as such. Before that, the internet was a different place. https:/…

> Gamergate or GamerGate (GG) was a loosely organized misogynistic online harassment campaign motivated by a right-wing backlash against feminism, diversity, and progressivism in video game culture Okay, what the actual fuck? IIRC it was people whining about the absolute state of games journalism in the 2010's.

GamerGate was about ethics in games journalism roughly as much as the Arab Spring was about a street vendor having his cart confiscated.

That was their initial spark, but it kicked off a ding-dong battle for years. You could argue it's still going today, given places like /v/ and ResetEra are still fighting it, games like Dustborn and Concord are pilloried, and the "Sweet Baby Inc. detected" Steam curator exists to list games that have taken that company's advice.

Re: 25 Years of Wikipedia

#193

Earlier quoted context omitted.

By now they should be sitting on a billion dollars that safely yields a forever self-funding annual income ($30m-$50m) that would pay for all of their necessary expenses. They would no longer require any donations. It's grotesque and wildly irresponsible how they're managing the organization. If LLMs become the center of knowledge resources going forward (which they will), Wikipedia's funding will decline as their tr…

LLM's can't just be "the center of knowledge" on their own, they need to learn and be trained if they are to be useful. A whole lot of LLM knowledge comes from Wikipedia to begin with.

statements on wikipedia are summarized from sources, LLMs once trained on wikipedia to summarize, can then summarize on their own from the source material, and probably with less bias

Re: 25 Years of Wikipedia

#194

Earlier quoted context omitted.

Indeed! Sometimes even more than actually exist! I don't think LLMs can be faulted on their enthusiasm for supplying references.

Yup, there's a wonderful, presumably LLM generated, response to somebody explaining how trademark law actually works, the LLM response insists that explanation was all wrong and cites several US law cases. Most of the cases don't exist, the rest aren't about trademark law or anywhere close. But the LLM isn't supposed to say truths, it's a stochastic parrot, it makes what looks most plausible as a response. "Five" is…

"Five" is not merely "plausible". It is the uniquely correct answer, and it is what the model produces because the training corpus overwhelmingly associates "2 + 3" with "5" in truthful contexts.

And the stochastic parrot framing has a real problem here: if the mechanism reliably produces correct outputs for a class of problems, dismissing it as "just plausibility" rather than computation becomes a philosophical stance rather than a technical critique. The model learned patterns that encode the mathematical relationship. Whether you call that "understanding" or "statistical correlation" is a definitional argument, not an empirical one.

The legal citation example sounds about right. It is a genuine failure mode. But arithmetic is precisely where LLMs tend to succeed (at small scales) because there is no ambiguity in the training signal.

Re: 25 Years of Wikipedia

#195
post #177

Earlier quoted context omitted.

> One thing I should clarify is that Wikipedia's purpose is to aggregate the current general view on things. So even if you know something is true personally, you cannot put it in unless you can find a reliable place where someone has documented it. In the cases I have there I had to first find the appropriate backing references before I could make something happen so it's not a trivial fix. This is where I would dis…

When you say you disagree, I assume you mean that you disagree that Wikipedia's approach is good. I don't think I was making that claim, however. I have no value position on Wikipedia's approach except that I appear to endorse it by participating. There are certainly true things that Wikipedia will not contain because they are insufficiently described in sources that Wikipedians find acceptable. But nonetheless that…

I meant that I disagree that Wikipedia is really trying to give a general view of events. That might have been the original intention, but it's not what it's doing in practice.

It does all hinge on that important list of acceptable vs unacceptable sources. In the last few decades there's been an increasing trend for news outlets to take a political position and decline to report on stories which would damage that position, which becomes most obvious whenever the US holds an election.

Re: 25 Years of Wikipedia

#196

Earlier quoted context omitted.

> Gamergate or GamerGate (GG) was a loosely organized misogynistic online harassment campaign motivated by a right-wing backlash against feminism, diversity, and progressivism in video game culture Okay, what the actual fuck? IIRC it was people whining about the absolute state of games journalism in the 2010's.

I was there, it was as Wikipedia describes it. Read the talk page. Edit: the replies to this comment demonstrate why this problem is intractable: people are very emotionally invested into their idea of how things unfolded, and outright reject other perspectives with little more than a "nuh-uh!".

I was there as well. It was absolutely not as Wikipedia describes it. If the claim was that some people participating in GG did so because they were sexist, fair enough. That was true and unavoidable because you get crazies in every group. But that was not some kind of universal thing, such that Wikipedia should be describing the movement unambiguously as "misogynistic".

Re: 25 Years of Wikipedia

#197
In the last 2 years, Wikipdia's quality declined. For me this was evident when they suddenly changed the UI. The new UI is more annoying. Perhaps it is nicer for Average Joe people, but for powerusers it is just annoying to use now. But this is not the only problem: many articles have a low quality, or they are so complicated that Average Joe doesn't understand them, which is ironic considering the UI was most likely changed to appease Average Joe.

I am also displeased with the constant pop-up or slide-in widgets. This is a general curse for browsers that ublock origin prevented. I hate this. My browser should not allow for any such slide-in banner. I am never interested in anything written there - usually it is a "gimme more money", but even if it is not, I simply don't CARE what is written on it. Even python used this, on their homepage, where they are even so cheeky that you can not fully disable this thing, unless you block it with ublock origin.

I feel that too many websites fail the user now. Wikipedia does too. The intrinsic quality is still better than the AI slop spam that Google amplified world wide, while also ruining Google search, but the quality used to be better in the past, on Wikipedia.

Re: 25 Years of Wikipedia

#198

[flagged]

> the persistent editorial and corporate bias and intellectual dishonesty

Musk is explicitly partisan and has repeatedly manipulated Grok's output to suit his agenda. How could you possibly consider Grok a worthwhile alternative if you take issue with intellectual dishonesty and corporate bias?

Re: 25 Years of Wikipedia

#199
post #32

I can't be the only one who feels that Wikipedia's quality has really started to go downhill over the past 5 or so years. I've noticed more and more articles which read as ridiculously partisan, usually around subjects with any link to politics or current events. That's probably linked to the increasing polarisation in the US, but I get the impression that the sites neutrality policies have gradually been chipped awa…

It’s more likely that you became more radicalized so what used to read as neutral seems partisan now.

I mean... this is a very real phenomenon, but probably not in the way you're thinking of.

There are many simple statements of fact that, 15 or 20 years ago, were as universally uncontroversial as "the sky is blue", but today are considered radically controversial political opinions, and will get you banned for most online platforms if you dare utter them.

Re: 25 Years of Wikipedia

#200

In the 2000s, in the tech world, the open source successes that were being talked about was always Apache and Linux. When Wikipedia started gaining a bit of traction, everyone made fun of it. It was the butt of jokes in all the prime time comedy shows. And I always felt like telling the critics - "Don't you see what is happening? People all over the world are adding their own bits of knowledge and creating this huge…

Wikipedia is surely a formidable source of knowledge, but > Something that goes beyond our daily vices of politics and religion. You are romanticizing. Wikipedia is a corporation, just like Work or University, and I personally assume anything corporate is being manipulated by the owners or the ruling oligarchy because they are structurally unreliable. This is especially veritable for Wikipedia. Create an account ther…

> Wikipedia is a corporation, and I personally assume anything corporate is being manipulated by the owners or the ruling oligarchy because they are structurally unreliable.

Many articles are fine. I don't think you can equate this to all of Wikipedia automatically.

Post reply on HN