How lazy are these people scraping Wikipedia that they can do a bit of research to see there’s an entire dump provided by Wikipedia directly?

https://dumps.wikimedia.org/

Their entire industry is built on not paying for data. They aren’t going to spend their own money optimizing this until they have to pay the costs for their design choices.

Spending money to make money is one of the pillars of capitalism. Getting things for free is anathema. Isn't it strange no one is calling out tech billionaires for "communism" yet the social safety net we need and pay for gets that ding on the daily?

I used to be a Wikipedia donor, until there was an article that made the rounds a few years ago that argued that Wikipedia-the-website is essentially funded in perpetuity, and that the "Jimmy Wales needs your help or kittens will die" scarebanners are actually raising money for the Wikimedia foundation to spend on its own political ends.

Somehow you have a problem donating to a free service that has imense utility, because of some random reasons someone posted online, but likely paying for tons of other crap that is proven to finance various stuff that you’d be against, but have no clue it’s happening.

The wikimedia foundation spends a ton of money on nonsense stuff. Look it up, a lot of people don't like the wikimedia foundation for very valid reasons.

The constant calls for donations are to support the foundation, not the website, but they make it look like Wikipedia could disappear at any moment.

Their cost for bandwidth is a rounding error vs their salaries and other costs that the foundation racks up.

https://youtu.be/MpeOFvxor_0

"its own political ends" being Wikipedia shaped things, that seems like a fairly good thing to donate to.

I’m personally cool with Wikimedia doing whatever they want with their notations since it’s public and everything, but the donation banners definitely lead you to believe that it goes directly towards hosting if you don’t dive in.

And that unless people donate right now, Wikipedia is actually at risk.

I've seen far too many comments of people who fell for that lie and donated even though they really didn't have money to spare, because the fear of losing such an important resource drove them to it.

I donated to "Wikipedia" for years until I learned that it was actually not really going to Wikipedia. I completely stopped now.

I found the Twitter thread that I believe was the source of this; claims that they granted money to culture war institutions:

https://x.com/echetus/status/1579776106034757633

I love Wikipedia as a free source of history, but I have no interest in donating money to them as a proxy in the culture war.

As far as I can tell “culture war” here means fighting racism, or fighting racism specifically in a way that you don’t like. Please correct me if I’m wrong.

> Please correct me if I’m wrong.

"culture war" in this context was me quoting a 3yo Twitter thread that made a lot of people, inc myself, realize that Wikipedia's "appeals" are deceptive. I was trying to get the comment up before the train left the station and I lost service. I didn't do further research as to what the organizations were.

I don't like the implication that by stopping my support of Wikimedia, I'm potentially racist/pro-racism. That doesn't feel intellectually honest. I was supporting Wikipedia because I think it's a valuable public resource. When I want to support a different goal, I give to an organization that specializes in it directly.

Bait-and-switch appeals make me feel like Wikimedia is not a responsible steward of its donations. It doesn't matter what they're giving the money to; it's not what they implied they needed it for.

Definitely understand the frustration around the bait and switch.

I was not trying to imply that by stopping your support for Wikipedia you’re potentially racist; rather that by referring to anti-racist organizations using the term “culture war” you are potentially racist.

Then we disagree on the value of Wikipedia. "History" has historically been written by those in positions of power and is biased by that. It feels very much in line with Wikipedia's value proposition of being a "free source of history" that they are trying to improve equity and unbias that source by adding more underrepresented voices. Funding racial equity programs seems a reasonable way of achieving that.

This thread (public link: https://nitter.net/echetus/status/1579776106034757633) mostly suggests that is where that funding is going. It also picks on a few specific cases and positions them as bad, but given that the thread is written from the standpoint that racial equity is bad, I'm not inclined to take the couple of cherry picked examples at face value. There are valid reasons to critique them.

I believe that the criticism is not about what the Wikimedia Foundation does specifically.

The criticism is that the Wikimedia Foundation makes people believe that Wikipedia needs funding, so that people donate thinking that they contribute to Wikipedia. Personally I have nothing against the Wikimedia Foundation, but when I learned that I was not actually donating to Wikipedia, I felt tricked. They lost my trust because they made me feel like I was tricked.

And my point is that for Wikipedia to be the best Wikipedia, these initiatives are important, and therefore I don't see them as that much different.

I know Wikipedia has a reputation for cringe donation drives, but at a basic level I don't disagree with them raising money, with what they're raising it for, or with most of how they pitch the donation requests, as continuing to maintain and improve Wikipedia. Focusing on the split between the Wikimedia Foundation (the org you donate to) and Wikipedia (the product they produce) is a bit of a red herring in my opinion.

Yeah that's fair. I'm just saying that they made me feel tricked, and apparently I'm not the only one.

What is Wikipedia by that measure?

The Wikimedia Foundation host Wikipedia. The servers, power, connectivity, bandwidth and engineers who maintain it are all paid from donations on the site.

But that's not Wikipedia.

Because the Wikimedia Foundation has Wikipedia, which I love, doesn't mean I love (or even know) everything the Wikimedia Foundation does. I would donate to Wikipedia, but they apparently don't need it.

Same thing for Firefox: I would donate to Firefox, but I have to give to Mozilla that does a bunch of other stuff and doesn't even treat Firefox the way I would want them to. So I don't donate.

It is not wikipedia shaping things. It's politics, travel and lunches.

The foundation does not csre about wikipedia. They dont even try to clean it up from the agents who "shape" articles. You very often can see a situation where two admins help each other. Nobody reviews this really. I mean you can ask a "check user" - who will decline to check if the admins arent sock puppet accounts / multi accounts.

Nobody from foundation fights the deletionist, dishonest admin clique and people who discourage everyone feom editing wikipedia by racing to have the most reverts.

(Also what is with the random quotes added to every article now, that look like writing of 12 year olds. No more concise articles)

Same. After learning that Wikipedia is not, in fact, about to go off the internet for lack of funds, I came to hold a very dim view Wikipedia organization.

The reason?

>> a rise they attribute to AI crawlers. AI companies are killing the open web by stealing visitors from the sources of information and making them pay for the privilege

Shouldn't be paying for bandwidth.

No such thing as a free lunch - whatever balance sheet it lands on, every byte of every request costs someone something.

Yep but Wikipedia exists to freely disseminate information to humans. Should they pay for the bandwidth for AI agents?

How to tell people you don’t know how carrier grade internet connections work.

Every request contributes to server costs. Every byte contributes to the 95/5 pricing. Wikipedia can’t afford private peering with every ISP in the world.

Bandwidth is dirt cheap. I do 30 petabytes per month.

I assume it’s for work, but man am I jealous. Cox Internet has me on a 1TB/mo cap.

No, the link exists, every second you saturate it or not, and if not it's just lost bytes.

That's not how bandwidth metering works at all.

Yeah, because they use Google Cloud. /s

If you think Wikipedia is delivering over a metered link, turn in your nerd card at the front desk.

I just about gave myself whiplash reading that, your critical thinking is all over the place and completely disconnected with reality.

You poorly assumed bandwidth is free (like many people do it is a common mistake), which tells me everything I need to know. Take your own advice: hand in your nerd card at the front desk, and don’t even think about passing go or collecting 200.

From 'What it Costs to Run Wikipedia'

Wikipedia spends $US 3m on hosting in 2024. They spend $US 107m on salaries.

So their hosting costs might have gone to, say, $US 5m at the outside.

That is they are maybe 1/20th of their salary cost.

AI's extra scraping might be a problem for some folks. But for wikipedia to talk about the percentage increase is disingenuous.

Source :

https://www.saastr.com/what-is-costs-to-run-wikipedia-3m-in-...

An increase in bandwidth costs is an increase. There's nothing disingenuous about that.

I imagine it also takes a toll on server, power, and maintenance (staff time) costs as well.

But a 50% increase of a relatively small amount is still a relatively small amount.

So it helps to show exactly how much more.

I don't understand the argument here. If your company had hosting costs increase this much and said "hey sien, we're going to cover the costs by cutting your salary" I don't think you'd be thrilled. Non-profits have staff and they have to pay them.

I think the fine print in a report is not an issue but the incendiary blue sky headline is.

> Non-profits have staff and they have to pay them.

Being the devil's advocate, there are are quite a lot of smaller non-profits that are entirely volunteer-run. While Wikipedia's mission might inherently be big enough to merit a few salaried employees, I can't fathom why they need so many. As per Wilde's famous quip, it seems to me that "the bureaucracy is expanding to meet the needs of the expanding bureaucracy".

> Being the devil's advocate

Honestly, just don’t. The world doesn’t really need hypothetical antagonism anymore.

From my experience, if no one is willing to act in hypothetical antagonism, you inevitably end up with toxic positivity.

So you think it's a fine trade for them to relinquish, say, 8-10 staff members in order to feed an AI machine they never asked for?

Do we even know what percentage of their hosting costs is bandwidth?

I'm curious what would happen if Wikipedia tried to sign up for a $200/month Cloudflare plan and use it for all their content. In theory, that would be consistent with Cloudflare's ToS; and while there have been examples of Cloudflare sales reps trying to pressure customers with larger bandwidth needs to spend more, executives have said (including in comments here on Hacker News) that those sales reps were acting against policy, and that unlimited bandwidth really does mean unlimited.