Amazon donates $1M to Wikimedia | Hacker News Reader

Amazon donates $1M to Wikimedia

techcrunch.com

337 pointsdoppp84 comments

"Apple, Facebook, Microsoft and Google all contributed around $50,000 by matching employee gifts."

That seems paltry for Google, given that their knowledge graph benefits a lot from Wikipedia content.

Google gave $2 million in 2010 from searches, and Brin gave $500k. I'm guessing they gave more at various times, but I think you're right, Wikipedia is a vital resource used by Google, Apple, Microsoft, and Amazon's AI agents, and they should be donating far more since it is arguably a significant value add for those products.

Throwing money at an organization has diminishing efficacy.

What do you think $2 million or $10 million or $100 million would have done for the Wikimedia organization, you think they would hide the donation <div> next year?

Establish an endowment to ensure they are not fighting for survival in the future.

In 2017 WMF had $120M in assets. It's probably closer to $150M today, since they are making over $20M a year in "profit" (revenues minus expenes). The cost of actually running Wikimedia projects (servers + network + salaries of dev ops and salaries of code maintainers) should easily fit below $10M, because it did in 2012. There has been growth since then in terms of usage, but this is very scalable growth that's cheap to handle: it's not servers and network expenses that's been skyrocketing, these are pretty much constant since 2012. It's the salaries, grants and awards. Look at their financial reports[1], and especially compare the growth in expenses over time. Suffice to say, if WMF struggles in the future, it's only due to mismanagement of its cancerous growth.

[1] - https://wikimediafoundation.org/about/financial-reports/

> It's the salaries, grants and awards.

Salaries for who?

Grants to who and for what?

Awards to who and for what?

Like, can someone provide a summary, for those of us who don't really have the time to dive into the reports?

They have around 300 FTEs. It might sound high, but they are the 5th most visited website in the world. Of others in the top 10, I think Reddit is the only other company with under 1,000 employees (at around 250). In comparison, Taobao which is also in the top 10 has over 30,000 employees.

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

https://meta.wikimedia.org/wiki/Wikimedia_Foundation_Annual_...

It was already top 8 in 2012, and had less than half of the headcount it has today, spending a third of what it spends today in salaries.

Eh - I find this public over-scrutiny of a company just because it's a nonprofit to be a little offputing. Sure, you want to know your donations aren't being wasted, but I keep in mind that most non-profits where my friends work have people under-staffed and without enough budgets overall compared to their parallels in commercial businesses.

Do you know for a fact there that what they were doing in 2012 was sustainable for their employees, and building a solid future path for growth? If everyone's overworked and not efficient, keeping it overly lean isn't good in the long term.

Sure, I was an admin on Wikipedia in 2008-2012, and we were doing completely fine. The Foundation was tenth of its current size at the time, and wasn't doing much except keeping the lights running, which is all the community needed. Bear in mind that Wikipedia is run by community, and if the Foundation has anything to say in Wikipedia's internal affairs, it is through employees who also happen to be community members. If the Foundation is overworked and not efficient, it literally doesn't matter, as they don't actually do much on Wikipedia itself.

The reports are the summary.

They could do that themselves and in fact they are - $5m/year goes into their endowment. Take a look at their budget:

https://meta.wikimedia.org/wiki/Wikimedia_Foundation_Annual_...

Quite a lot of waste in there, e.g. $850k on "brand strategy" - kind of looks like they have more money than they know what to do with.

https://en.wikipedia.org/wiki/Wikipedia:Wikipedia_Signpost/2...

I never saw these "Signposts" before, interesting. Was there ever an opposing view released to this, uh, blogpost? Signpost?

edit: apparently there is a kind of discussion section at the bottom

WMF has been throwing out money for virtually everything except the actual *pedias.

Whatever it takes not to see Jimmy Wales' sad puppy dog mug begging for $5 donations at the top of every page.

I think no matter how much you donate to Wikimedia, they still like having the sad puppy dog. Wikimedia has a huge endowment, with the way the stock market did in recent years, it has grown significantly too.

Your money could have more impact elsewhere.

You can move to Finland, they don't ask for donations here.

In fact, trying to donate gets you a "Thank you for your interest" page: https://donate.wikimedia.org/w/index.php?title=Special:Landi...

The legislation could use modernization, but progress is slow...

> by matching employee gifts

The employees of those companies collectively donated $50,000.

Also, WMF has more than enough money to run Wikipedia probably indefinitely anyway. They do other things, but Wikipedia is at no risk of going away.

That seems typical for Google. Since when do they pay content creators properly?

This is called putting lipstick on the pig.

The pig, in this case, being marginal utility theory.

Amazon should be picking up the tab on the AWS costs. Great PR move, great for customer case studies etc and slash what is probably 2nd or 3rd biggest cost driver.

Useless signaling. Wikipedia definitely doesn’t need any more money. This $1M will be burned on salaries of Wikimedia Foundation employees, majority of which do nothing useful. Look at their financial statements and see for yourself. Wikimedia has enough cash already to find operations for a decade or two, and they spend most of it on salaries of people running programs you never hear about, and which have no bearing on the way most people actually use Wikipedia and Commons.

> Useless signaling.

Their contribution to Wikimedia sends a positive signal to potential customers that is very possibly less useless than the more typical form of corporate signalling: advertising.

Do you think that advertising is also useless signalling?

Useless from the point of view of the recipient. I believe GP meant "useless" as in "PR stunt"; certainly not useless for their public perception, unfortunately.

Useless for Wikipedia readers, not for the donor nor for the Wikimedia foundation members who will spend the money.

I think advertising reaches way more people and generates income, thus isn't useless.

Wikimedia salaries: https://meta.m.wikimedia.org/wiki/Wikimedia_Foundation_salar...

These are only some of the salaries of a small fraction of the employees for which the salary number is even avaiable. WMF employs 5 or 6 times as many people, and it’s not clear to me what half of them are even doing. I was an admin on Wikipedia in 2007-2010, when WMF was tenth of its current size, and the Wikipedia community never felt the lack of manpower on the foundation side. Now they are an order of magnitude bigger, but in terms of end to end impact it hardly changed at all.

It's the salaries disclosed in their form 990, so yes, not all salaries.

Here's the 990 for 2016-2017: https://upload.wikimedia.org/wikipedia/foundation/6/67/Form_...

Those salaries seem low, are they not based in SF?

They are, but they are a non-profit, not a VC backed unicorn.

Interesting. Does this page show that they have had a CTO for just 3 of the past 10 years? Or only that they paid a CTO for those years?

Not sure. They announced Victoria Coleman as their CTO in 2016[1]. But she isn't listed in their 2017-2018 Form 990[2]. Odd, since all "company officers" are supposed to be listed, even if paid nothing[3].

[1]https://blog.wikimedia.org/2016/11/02/victoria-coleman-chief...

[2]https://upload.wikimedia.org/wikipedia/foundation/6/67/Form_...

[3]https://www.irs.gov/charities-non-profits/form-990-part-vii-...

Useless signaling? 1 million is definitely not useless.

It is absolutely useless to Wikimedia, since it already rakes in around $90M a year, and only manages to spend $70M of it, most of it on useless projects that don't make any difference to actual day to day operations of Wikipedia, which is done 95% by unpaid volunteers (the remaining 5% is paying for servers and network, and maintaining software, which costs much, much less than $70M).

Wikimedia clearly doesn't need as much money, but since it gets it, it grows to spend as much of it as it can. If they scaled back to their spending in 2012 (when they already were high up in the top 10 most visited websites), and did fundraising for only extra 2 or 3 years, they would have enough endowment to fund their operations in perpetuity, and not need to run sad Jimbo face on Wikipedia ever again. Alas, the foundation will expand to fill the available space, wasting millions in the process.

WMF is essentially in the agency-budgeting phase. They always need a larger budget and they always find a way to (almost) blow through their previous budget to have an excuse to ask for a bigger one next year round.

> We've been selling cocaine for the CIA so they could buy arms from Iran? Did we at least free some hostages?

> No, this was about our annual budget.

> What?

> If we don't spend it this year we can't get an increase next year.

Google went after Freebase, did all the knowledge transfer to their graph and discontinued the project. Wikipedia now became a vitally important source of information and it is great to see big players trying to keep open knowledge base funded.

Wikidata is the new better Freebase:

https://www.wikidata.org/wiki/Wikidata:Main_Page

https://www.wikidata.org/wiki/Q2 'Earth' example

I'm sure this has been discussed before on HN, but how has wikimedia avoided the fake news problem that plagues facebook, twitter and youtube.

Wikipedia definitely is not immune to the fake news effect. I can recall both reading about and directly seeing numerous instances of propaganda-esque edits. I'm guessing it probably just seems less prevalent because readership isn't concentrated on any one Wiki article the way that news/fake news is. I'm also under the impression that most wiki articles are closely watched by their authors as well as various wiki bots that alert to certain types of changes, so many/most malicious edits are probably noticed and reverted soon after they are made.

Wikipedia has started to become _dare I say_ hostile to outside editors. The current glacial pace that it is updated at maintains pace with science, and journalism. But rarely the larger internet.

While this is frustrating, when your trying to correct a citation or trivial spelling error in a long forgotten page. While on frequently trafficked pages this pace prevents a lot of malicious behavior.

My Wikipedia account was created two months ago and has exactly two edits, one of which removed a spam link and another which fixed a transliteration error. There has been no hostile reaction.

If you're just trying to correct a citation or trivial spelling error on a long forgotten page, editing doesn't seem to be particularly difficult.

You don't even need an account to do those changes.

I think this goes to show Wikipedia generally has its priorities in order. Although I'm sure the process could be improved, friction against trivial corrections is a small price to pay for effectively defending the integrity of content against psychological warfare and vandalism, IMHO :)

They have had the fake news problem for much longer and have multiple editors who have pet topics that they will maintain to their liking. "Fake news" only exists as a function of the design of those websites, all social media relies upon feeding the ad beast, which means more usage. Since it is fundamentally different in creation it does not have the same issues but has different issues.

https://en.wikipedia.org/wiki/Criticism_of_Wikipedia#Exposur...

There is a lot more direct human intervention/judgement on Wikipedia, there are very clear standards for citation, evidence, and writing, and there are mechanisms for identifying and removing troll submissions while including explicit justification for why it was removed. Wikipedia was built from the ground up to oppose bad actors in ways that are constitutionally impossible for advertising-monetized platforms.

Wikipedia is very well cited. You can easily find counterexamples too, but pretty much all "long" articles that's been reviewed by editors are well cited. Facebook is never cited, I've literally never seen anyone citing source on facebook. Polar opposites.

Oh, they have it.

Unfortunately, some people get tunnel vision about fighting it, and then the propagandists use social media to circulate dishonest smears about them that spread widely out of context. There was an example right here on HN: https://news.ycombinator.com/item?id=17109290

Edit: In fact, the propaganda operation is still ongoing: https://news.ycombinator.com/item?id=18051210

People don't need reliable citations to post on FB.

There's also the visibility thing. You can put fake news on FB without making it visible to everyone.

Most fake news isn't about spreading disinformation for the sake of corrupting a population. It's about generating clicks and finding suckers. Wikipedia isn't a great platform for that.

Considering all of their 'Alexa what is [x]' information comes from wiki, this seems short-changed.

It’s kinda gross how little the big tech players have donated considering how much of their and academia’s AI efforts leverage wikimedia data.

As other commentators have said, this donation will definitely invite accusations of virtue signaling. What Amazon needs to do is pay (not donate!) its employees better but that's also stating the obvious.

What intrigues me more is that each era's barons find certain causes more appealing than others. Carnegie gave a lot of money to public libraries at a time when the printed word was expanding its reach. It's not surprising that Wikipedia is a favorite amongst tech tycoons along with education.

Amazon has 500K+ employees. $1M is what, two bucks per employee? Maybe three bucks if we exclude higher management, they get enough anyway. So you're saying "instead of donating to Wikipedia they should give each of their employees a free Starbucks coffee once a year". That would surely change things.

You need to read what I wrote more carefully. I said "Amazon needs to pay its employees better" as in a living wage - not distribute the one mil it gave Wikipedia amongst its employees.

Why would you mention it as a comment to Amazon donating to Wikipedia unless you see some link between this $1M and Amazon workers pay?

Nice to see good news like this.

That's a good point. This is good news. Why don't people celebrate good news?

Because it is neutral news. Wikimedia has way more money than they can burn any time soon, and money has diminishing returns.

at this point in time in the dumpster fire, neutral should be celebrated!

I don't understand this sort of pessimism. If a news is good, we should say good. If a news is bad, we should say bad. If it's neither, then we should state it's neither. What makes you think we live in an especially bad time frame?

the assault on truth, misinformation and disinformation.

the rise of authoritarianism.

the polarization of poltics.

climate change denial.

jerks getting outraged when people suggest not being a jerk.

rage-baiting for clicks.

How is this random potpourri of global problems supposed to prove your point? Do we have any data we have more problems now than before? According to what basis? Even if this is true, what change does this make, why does this imply we should start praising neutral news just because it's not bad.

a potpourri it may be but it is a response to your question about why I think this is an especially bad time frame. I see no reason not to consider them relevant and I also see them as being indicators that things could be a lot better. Based on this my original comment was a desire to try to increase positivity.

Waste of money. Wikimedia is still trying to support the ancient and effectively deprecated Mediawiki which is only realistically useable for corporate style deployments. Individuals cannot use it well and wikimedia fails at meeting its mission statement time and time again. Donations to wikimedia are effectively donations to wikipedia, which cannot house the worlds information properly. Information craves distribution and perspectives, decentralization.

I want to see proffessors and trade skill teachers create good wiki style sites, not for yet more useless projects by wikimedia designed to bolster their salaries.

It'd be possible to make an absolutely free hosting site for informational wiki's (a wiki farm) for 1M that'd last a generation. But nobody really cares about that.

What exactly is the problem with MediaWiki? I for one don't believe that any problems Wikipedia currently has are of technical nature. Lots of people has contributed greatly to Wikipedia in 2000s, when MediaWiki was even less accessible than it is right now.

Setting up MediaWiki properly yourself is not a small undertaking. How wikipedia does what they do is with puppet deployments and a significant amount of automatic server management. It's incredibly inefficient given that pages should be statically served, it's very difficult to even know what to install and it may not even be future compatible.

For example getting VisualEditor, MediaWiki's custom PHP-lua extension for your version of PHP, needing PHP-FPM, managing versions of PHP, correct configs. Setting up page caching, knowing what extensions to get, seeing lots of out of date ones, having to configure them properly. Ontop of all this doing it actually right with proper file and user permissions is just completely inaccessible to anyone but those who make a job of it.

Even the mediawiki 'package' is out of date and works differntly, causing more problems.

It should be as simple as getting a domain, getting cloudflare, pop up a $5 VPS and a couple commands to have a proper and automatically updated install. User sets up website, sets backups to GDrive or some other cloud service, allows logins from google, facebook, whatever, and can then manage everything from the UI.

Right, these are all real problems with the software, but then again, none of these is really a problem for Wikipedia itself at all. For Wikipedia, all of the problems you mention have been already solved, everything is set up and functional.

And that's the problem. The world's information cannot be in one place, and mediawiki is made to look like it's 'for information' when in reality it's for Wikipedia. It has intentions to be extensible but fails to follow through, and this is incredibly misleading.

The formats used to store data are an incredible mess. https://github.com/spencermountain/wtf_wikipedia has a take on it.

This is the equivalent of me announcing a $3 donation to Wikipedia. PR signaling and nothing more.

Throwing money at Wikimedia foundation doesn't solve a problem either.

Typically the reason why wealthy individuals and organizations opt to do "matching" is because there isn't a $number that will satisfy people or even help the situation.

So then you could tie your contribution directly to public interest.

The world is open to better ideas, so try to contribute those.

Large companies donating large sums of money non-anonymously seems to circumvent the 'we're neutral because we don't have sponsors' credo.

They'll get used to the large cash injections, that is how political power is bought.

Wikipedia already has a problem with cash injections and not being able to save extra money instead of launching more and more projects.

As for the point you're making, I disagree as long as the income from one company isn't reaching too big a percentage of their total yearly income. And then again, Firefox clearly wasn't politically subservient to Google, despite most of their income coming from there.

So it's not so much an automatic reaction, the internal culture plays a big role. Which is why Wikipedia pre-existing issue with incoming cash usage is a red flag.

good - they deserve it!

Chump change for Bezos. Really.

It’s Amazon the company not Bezos the person.

At this point, can you really separate the two? Most of Bezos’s wealth is tied up in Amazon, and he is virtually in control of the company.

According to Bloomberg [1] he owns "just" 16.4 percent.

Sure, he is CEO and probably has a large influence on the board too.

But just based on stock, he doesn't even seem to have a blocking minority. (Unless he has a lot of preferred shares of some kind).

He's the shot-caller. Nobody else has that much influence in the company.

Well it's not directly out of his pockets. It might impact this year's bottom line, but he didn't paid the 1M directly.

> Chump change for Bezos. Really.

Ok, then how much should he have given? Keep in mind that Wikipedia didn't even need the money, because they have enough already.

Reply on news.ycombinator.com