Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Wikimedia Enterprise – APIs for LLMs, AI Training, and More

enterprise.wikimedia.com

Post reply on HN
Print view View original

131–140 of 166 posts

Re: Wikimedia Enterprise – APIs for LLMs, AI Training, and More

#131

Post by admissionsguy »

bawolff wrote: post #23

While i am sympathetic to wmf finding alternative funding streams, I do get nervous about these sort of things due to the inherent conflict of interest and incentives to canabalize the free offerings. I'm not saying that is happening now, but will it happen eventually? Additionally, originally it was promised this would all be open source, and officially they are sticking with that, but it seems like they are going w…

WMF goes the way of bureaucracies. It finds multiple ways to enlarge itself over the years and just cannot ever get any smaller. I suppose at some point it will overextend itself and harm Wikipedia, the only[1] thing of value, in the process.

[1] Yeah there are Wikidata and some other projects. But it's like saying Google is not an ad company.

Re: Wikimedia Enterprise – APIs for LLMs, AI Training, and More

#132

Post by mrweasel »

bhickey wrote: post #117
flipbrad wrote: post #113

Earlier quoted context omitted.

It would be nice if we had a "lot of lawyers", given how frequently we're sued to try and get content censored, or having to fight orders to hand over user data - and more generally, how massive these new laws we need to comply with are (see, e.g., the EU Digital Services Act, which even creates an entirely new annual independent audit process). We even intervene in other court cases to try and prevent bad laws being…

It isn't a question of the good work you do. People care about Wikipedia, not the Wikimedia Foundation. The criticism arises from misleading advertising. WMF fundraising conflates the two, implying that _Wikipedia_ needs money or it'll die. Meanwhile the 2023 budget shows $3.1m in hosting expenses versus $24.4m in awards and grants.

That is an issue. There is a number of projects that the Wikimedia Foundation want to do or be involved in, because they align with the mission. These all costs money, but are frequently of little interest to anyone not involved directly. There is absolutely no way to fund these, which leads to the foundation pushing for donations via Wikipedia, because that's the only thing enough people actually care about.

For the most part Wikimedia could kill off everything but English, Germany, French, Russian and a handful of other wikis and most people would be just as happy.

Wikimedia absolutely suck at telling people why they need the money. Technically the budget is completely transparent, it's just communicated extremely poorly.

Re: Wikimedia Enterprise – APIs for LLMs, AI Training, and More

#133

Post by chuckreynolds »

miffy900 wrote: post #120

I'm reading the API docs https://enterprise.wikimedia.com/docs/ And they don't have an OpenAPI spec available to download? So they seriously expect developers to just manually write their own client code by reading and translating those example CURL commands by hand??! Seriously it's 2024! Not having a spec to download and insepct for any API is a sign of incompetence. When tools like Postman or https://github.com/Op…

link at bottom of https://enterprise.wikimedia.com/docs/ page

Re: Wikimedia Enterprise – APIs for LLMs, AI Training, and More

#134

Post by schoenobates »

miffy900 wrote: post #120

I'm reading the API docs https://enterprise.wikimedia.com/docs/ And they don't have an OpenAPI spec available to download? So they seriously expect developers to just manually write their own client code by reading and translating those example CURL commands by hand??! Seriously it's 2024! Not having a spec to download and insepct for any API is a sign of incompetence. When tools like Postman or https://github.com/Op…

Bottom of the page ...

https://api.enterprise.wikimedia.com/spec/spec.yaml

Re: Wikimedia Enterprise – APIs for LLMs, AI Training, and More

#135

Post by JimDabell »

yareal wrote: post #91
tw04 wrote: post #76

Earlier quoted context omitted.

What are you talking about? The AVERAGE CD right now is 5%. My local CU is almost 6%. US bonds are currently ~4.5% - if you consider those unstable, I guess the US economy isn't stable - and if the US economy crashes, wikipedia will be the least of their or our worries. Wikimedia's expenses are almost ENTIRELY going to staff. Their balance sheet for 2023 included $101m in expenses for salaries and benefits out of a t…

An engineer costs $500k a year. Salary, benefits, office space, equipment, hr, legal, and other overhead. The engineer will only see a fraction of that, of course. If you told me it took a hundred engineers to run Wikipedia I'd say, that's not totally unreasonable. Features, design, api, scaling, moderation, there's a ton for engineers to be doing.

An engineer doesn’t cost $500k/yr. An engineer who lives in one of the highest cost-of-living places on earth costs $500k/yr. There’s absolutely no reason Wikimedia needs to pay that much.

Re: Wikimedia Enterprise – APIs for LLMs, AI Training, and More

#136

Post by krainboltgreene »

miffy900 wrote: post #120

I'm reading the API docs https://enterprise.wikimedia.com/docs/ And they don't have an OpenAPI spec available to download? So they seriously expect developers to just manually write their own client code by reading and translating those example CURL commands by hand??! Seriously it's 2024! Not having a spec to download and insepct for any API is a sign of incompetence. When tools like Postman or https://github.com/Op…

> And they don't have an OpenAPI spec available to download? So they seriously expect developers to just manually write their own client code by reading and translating those example CURL commands by hand??!

This is baby work to literally every programmer I know.

Re: Wikimedia Enterprise – APIs for LLMs, AI Training, and More

#137

Post by flipbrad »

bhickey wrote: post #117
flipbrad wrote: post #113

Earlier quoted context omitted.

It would be nice if we had a "lot of lawyers", given how frequently we're sued to try and get content censored, or having to fight orders to hand over user data - and more generally, how massive these new laws we need to comply with are (see, e.g., the EU Digital Services Act, which even creates an entirely new annual independent audit process). We even intervene in other court cases to try and prevent bad laws being…

It isn't a question of the good work you do. People care about Wikipedia, not the Wikimedia Foundation. The criticism arises from misleading advertising. WMF fundraising conflates the two, implying that _Wikipedia_ needs money or it'll die. Meanwhile the 2023 budget shows $3.1m in hosting expenses versus $24.4m in awards and grants.

Firstly, there's less conflation these days - go see recent banner wording for yourself. Secondly, if you're still just acknowledging Wikipedia hosting costs - and thus pretending there's (for example) no legal work necessary for it - I don't think people are getting through to you as they should. (And no, I'm not saying all legal work we do is a strict necessity for Wikipedia. Some is a strict necessity, and some is strategic e.g. an amicus, or the NSA lawsuit - but the latter does help secure a healthy environment for it and future projects that might want to take its place.)

Re: Wikimedia Enterprise – APIs for LLMs, AI Training, and More

#138

Post by causal »

krainboltgreene wrote: post #136
miffy900 wrote: post #120

I'm reading the API docs https://enterprise.wikimedia.com/docs/ And they don't have an OpenAPI spec available to download? So they seriously expect developers to just manually write their own client code by reading and translating those example CURL commands by hand??! Seriously it's 2024! Not having a spec to download and insepct for any API is a sign of incompetence. When tools like Postman or https://github.com/Op…

> And they don't have an OpenAPI spec available to download? So they seriously expect developers to just manually write their own client code by reading and translating those example CURL commands by hand??! This is baby work to literally every programmer I know.

Not to mention that if it's really needed, the LLMs this is made for also excel at converting CURL commands to whatever spec pretty trivially. I've used them this way on a few occasions.

Re: Wikimedia Enterprise – APIs for LLMs, AI Training, and More

#140

Post by cdme »

riku_iki wrote: post #57
cdme wrote: post #50

Earlier quoted context omitted.

No, but nobody is obligated to contribute to or support them either.

that's not really true. Because of network capture, those who wants to contribute into some public free human knowledge, obligated to do it through wikimedia and contribute to wikimedia revenue streams.

Or this simply discourages contributions for folks not wanting to volunteer effort to the gaping maw of yet another monolithic content vacuum/LLM.
Post reply on HN