"Overt racism" is a better way to phrase it, I suppose. That early boom of modern social media in the early '10's at least had people use dogwhistles or anonymity. Saying half the stuff today...
"Overt racism" is a better way to phrase it, I suppose. That early boom of modern social media in the early '10's at least had people use dogwhistles or anonymity. Saying half the stuff today would have ended political careers.
The problem is that when you create parallel products, being "Foo but conservative" isn't a selling point by itself except to preach to the choir. The only way to kill Wikipedia is to subvert it...
The problem is that when you create parallel products, being "Foo but conservative" isn't a selling point by itself except to preach to the choir.
The only way to kill Wikipedia is to subvert it directly, then coast on the branding. Elon should know this, having done this exact strategy with Twitter.
The fact that he's doing this instead tells me he's out of ideas and is trying to grab headlines by shouting "AI" at the top of his lungs.
" The Shadow that bred them can only mock, it cannot make: not real new things of its own. I don't think it gave life to the orcs, it only ruined them and twisted them; "
" The Shadow that bred them can only mock, it cannot make: not real new things of its own. I don't think it gave life to the orcs, it only ruined them and twisted them; "
I don't know if he ever made any offers behind the scenes but Wikipedia has outright said they aren't for sale in response to speculation that Musk wanted to buy it one of the previous times he...
I don't know if he ever made any offers behind the scenes but Wikipedia has outright said they aren't for sale in response to speculation that Musk wanted to buy it one of the previous times he complained about it.
But someone needs to fill it. When it comes to those contributors, trying to be factual at any level invites some critical thinking. Otherwise it turns into a memepedia which isn't something...
But someone needs to fill it. When it comes to those contributors, trying to be factual at any level invites some critical thinking. Otherwise it turns into a memepedia which isn't something people will take seriously (I hope).
Or maybe Elon just shoves it full of AI garbage. (It would be pretty interesting how that would be received too.)
It does indeed sound like he plans to let his LLM read every Wikipedia article, copy it, and rewrite it. Bias aside, it's going to be basically useless because of hallucinations, but it will...
“Grok is using heavy amounts of inference compute to look at, as an example, a Wikipedia page, what is true, partially true, or false, or missing in this page,” he said. "Now rewrite the page to correct, remove the falsehoods, correct the half-truths, and add the missing context.”
It does indeed sound like he plans to let his LLM read every Wikipedia article, copy it, and rewrite it. Bias aside, it's going to be basically useless because of hallucinations, but it will probably gain some popularity among his fans.
AI lawsuits are complicated, but that sounds to me like a prime example of one that could win. The precedent so far seems to be that it's fine to gobble up everyone else's data, regardless of if...
AI lawsuits are complicated, but that sounds to me like a prime example of one that could win. The precedent so far seems to be that it's fine to gobble up everyone else's data, regardless of if you would normally have the right to even have it, for as long as your product transforms what it is. For example, an LLM is not a book.
Saying that they want to hoover up someone's data to repackage it as the exact same kind of directly competing product though is an entirely different beast. If Amazon was ingesting books with the purpose to emit books that are direct replacements to existing ones so that they could keep all the sale proceeds I think those authors would have very strong cases. So unless Wikipedia's licensing allows for that it seems very legally problematic.
Wikipedia articles are licensed under the CC BY-SA license, which (not a lawyer) appears to be compatible with what he's suggesting to do, provided there is a link to the original page, a clear...
Wikipedia articles are licensed under the CC BY-SA license, which (not a lawyer) appears to be compatible with what he's suggesting to do, provided there is a link to the original page, a clear indication that the original has been modified, and that the content is re-licensed under the same conditions as the original.
The license wasn't made with generative AI in mind, but the idea of basing a competing encyclopedia on modified versions of Wikipedia articles isn't inherently against the spirit of the project nor the letter of the license, provided it's done in the right way. In fact, Wikimedia Enterprise offers for-pay APIs to make it easier for corporations to re-use their content, but the for pay bit is explicitly not about paying for the right to reuse the content. It's for accessing specific APIs to make it easier to extract, work with and keep up with changes in Wikimedia content, as well as customer support for those APIs and services. The content itself is free, and Wikipedia explicitly tells corporate users not to ask them for permission to reproduce content, as that right is not Wikipedia's (it's covered by the license and the content belongs to the individual contributors, from a copyright standpoint Wikimedia owns almost none of its own content).
Okayy, I assumed the point of the Wikipedia clone was to provide Grok "unwoke" data feast upon. Isn't there an obvious problem with regression if he's still planning on doing that?
Okayy, I assumed the point of the Wikipedia clone was to provide Grok "unwoke" data feast upon. Isn't there an obvious problem with regression if he's still planning on doing that?
What's wrong with good old conservapedia? Oh right, it doesn't have AI. Always best to have inaccuracies put in to your knowledge base at random, on purpose.
What's wrong with good old conservapedia?
Oh right, it doesn't have AI. Always best to have inaccuracies put in to your knowledge base at random, on purpose.
Some Biblical scholars have calculated that the earth is around 6000 years old, using the genealogical information provided in Genesis. Some biblically illiterate scientists, however, disagree with this method of calculation, falsely predicting it to be 4.5 billion years old.
I like that the "four angels standing on the four corners of the earth" as evidence for a flat earth is dismissed by "being a figure of speech", yet the same logic cannot be applied to the...
I like that the "four angels standing on the four corners of the earth" as evidence for a flat earth is dismissed by "being a figure of speech", yet the same logic cannot be applied to the creation story in Genesis, that each "day" is a figure of speech.
Cherry picking season is over, go home Conservapedia.
My bigger takeaway is relief that "Conservapedia" is an actual website and not some derogatory name for Wikipedia. I was worried for a second that Wikipedia had some scandals and shifting...
My bigger takeaway is relief that "Conservapedia" is an actual website and not some derogatory name for Wikipedia. I was worried for a second that Wikipedia had some scandals and shifting ideologies/political slants I wasn't aware of. (Which, it almost certainly does, but the problems would be way worse than I thought if it got casually called that.)
The wikimedia foundation who owns wikipedia has some very liberal leaning practices, but AFAIK Wikipedia itself doesn't get involved with those. My only gripe with Wikipedia is how often they beg...
The wikimedia foundation who owns wikipedia has some very liberal leaning practices, but AFAIK Wikipedia itself doesn't get involved with those. My only gripe with Wikipedia is how often they beg for donations like the site is about to shut down when the company generates insane amounts of revenue and the vast majority of their money goes to doing stuff other than running the site. That and nearly ALL of the actual work done on Wikipedia is by unpaid volunteers, so it feels a little greedy to solicit so much in donations when you're relying on free labor to run your site. (This is almost entirely because I'm jaded from being an unpaid volunteer reddit mod for a decade and watching them turn into a hundred billion dollar company off the work I did for them while simultaneously ruining the website and spitting in the mods faces time after time.)
I love Wikipedia, but whenever I get a popup to donate, it serves as a nice reminder for me to donate to the Internet Archive instead.
Exemplary
Wikipedia is how often they beg for donations like the site is about to shut down when the company generates insane amounts of revenue and the vast majority of their money goes to doing stuff other than running the site.
I love Wikipedia, but whenever I get a popup to donate, it serves as a nice reminder for me to donate to the Internet Archive instead.
Same, Wikipedia is an invaluable resource supported by thousands of incredibly hard working and dedicated people, but it also isn't in any great need. There are far more deserving causes out there...
Same, Wikipedia is an invaluable resource supported by thousands of incredibly hard working and dedicated people, but it also isn't in any great need. There are far more deserving causes out there for you to donate to, the Internet Archive is a great example.
It's not like they are hiding their financial reports The comments made by @OBLIVIATOR appear to be largely incorrect. It looks like Wikimedia makes money mostly from donations and endowments,...
It looks like Wikimedia makes money mostly from donations and endowments, spends a large proportion of it on hosting and technology, gives out some to support it's volunteers, spends some on admin and advertising and invests the rest because they're planning/hoping to be around a long time. The only thing that seems "insane" to me about the money they make is how they managing to run one of the world's biggest websites on so little money, frankly. They were operating on less than $180m in revenue last year!
A friend of mine is a DevOps guy at wikimedia. He's not well paid by the standards of his industry, he could easily make way more working for one of the big bad tech firms but he really believes in what they're doing and I have a lot of respect for that.
I don’t know anything about how much money Wikipedia has or how it spends it, but just reading the political tea leaves tells me it would be a good idea for them to be building up a rainy-day war...
I don’t know anything about how much money Wikipedia has or how it spends it, but just reading the political tea leaves tells me it would be a good idea for them to be building up a rainy-day war chest. For legal defense and technical defense alike. The Internet Archive needs this too, more urgently, as far as I can tell.
No worries, he and his minions can use Grok to just vibe code it. And for some weird reason, variables and functions and such will reference or outright use nazi names and imagery. Just for the lolz.
No worries, he and his minions can use Grok to just vibe code it. And for some weird reason, variables and functions and such will reference or outright use nazi names and imagery. Just for the lolz.
Is he going to vibe-code an operating system and office suite?
Last month, the billionaire also tweeted his plan to take on Microsoft by creating a new business called “Macrohard” dedicated to releasing rival software products with the help of AI.
Is he going to vibe-code an operating system and office suite?
ok elon aside, I would kinda like an AI encyclopedia, as long as it cites the sources where it got the info from and is somewhat reliable, and all articles are clearly labelled as AI generated....
ok elon aside, I would kinda like an AI encyclopedia, as long as it cites the sources where it got the info from and is somewhat reliable, and all articles are clearly labelled as AI generated. perplexity kinda tried but their feed is only composed of news so not very interesting
I'd be concerned about hallucinations but mostly I just don't see the value. I'm curious to where you're coming from though, what advantages do you see an AI encyclopedia offering over community...
I'd be concerned about hallucinations but mostly I just don't see the value. I'm curious to where you're coming from though, what advantages do you see an AI encyclopedia offering over community curated resources like Wikipedia?
it doesn't have to be an alternative to wikipedia, it can be it's own thing. the really cool thing about wikipedia is that the community can compile all the relevant details from different sources...
it doesn't have to be an alternative to wikipedia, it can be it's own thing. the really cool thing about wikipedia is that the community can compile all the relevant details from different sources into one article. if AI can do it, it'd be nice. because a lot of the time, the information you need is scattered around different web pages, articles, comments on reddit etc. and about hallucinations, imo perplexity has reduced the hallucinations quite a bit, I think it's reliable enough. but again, it doesn't have to be a replacement for content written by humans, just sort of something to exist alongside it. and in fact, I think it'd be pretty efficient if not everyone is generating content on it, like imagine if used the best LLMs available now (e.g. GPT 5 with the highest reasoning effort with mixture of agents). basically quality over quantity. if done right, it could actually be higher quality than some of the "educational" websites
Surprised he isn’t calling it Xipedia, he is obsessed with X after all. Is it because “xpedia” will sound too much like “Expedia”, which is an actual company?
Surprised he isn’t calling it Xipedia, he is obsessed with X after all. Is it because “xpedia” will sound too much like “Expedia”, which is an actual company?
Cool, AI powered pseudo-scientific racism backed by unlimited money
what could go wrong?
I miss that brief window in time where racism wasn't profitable. But I suppose Lyndon B Johnson's quote made this inevitable.
We had one of those?
"Overt racism" is a better way to phrase it, I suppose. That early boom of modern social media in the early '10's at least had people use dogwhistles or anonymity. Saying half the stuff today would have ended political careers.
of course, 2016 took a very hard turn.
Gotcha with the mention of LBJ, I had a much earlier timeframe in mind. Sorry for the confusion
I'd bet against it succeeding. Many people have tried to build Wikipedia clones, but none have gotten traction so far.
But what if you want a Wikipedia that is controlled by one billionaire asshole instead of a community of people striving to share knowledge?
The problem is that when you create parallel products, being "Foo but conservative" isn't a selling point by itself except to preach to the choir.
The only way to kill Wikipedia is to subvert it directly, then coast on the branding. Elon should know this, having done this exact strategy with Twitter.
The fact that he's doing this instead tells me he's out of ideas and is trying to grab headlines by shouting "AI" at the top of his lungs.
" The Shadow that bred them can only mock, it cannot make: not real new things of its own. I don't think it gave life to the orcs, it only ruined them and twisted them; "
He probably already tried to buy it.
I don't know if he ever made any offers behind the scenes but Wikipedia has outright said they aren't for sale in response to speculation that Musk wanted to buy it one of the previous times he complained about it.
What many people want is a Wikipedia that produces facts that align with their feelings.
But someone needs to fill it. When it comes to those contributors, trying to be factual at any level invites some critical thinking. Otherwise it turns into a memepedia which isn't something people will take seriously (I hope).
Or maybe Elon just shoves it full of AI garbage. (It would be pretty interesting how that would be received too.)
It does indeed sound like he plans to let his LLM read every Wikipedia article, copy it, and rewrite it. Bias aside, it's going to be basically useless because of hallucinations, but it will probably gain some popularity among his fans.
There's going to be a cottage industry of YouTubers finding ridiculous hallucinations in articles and making fun of them
AI lawsuits are complicated, but that sounds to me like a prime example of one that could win. The precedent so far seems to be that it's fine to gobble up everyone else's data, regardless of if you would normally have the right to even have it, for as long as your product transforms what it is. For example, an LLM is not a book.
Saying that they want to hoover up someone's data to repackage it as the exact same kind of directly competing product though is an entirely different beast. If Amazon was ingesting books with the purpose to emit books that are direct replacements to existing ones so that they could keep all the sale proceeds I think those authors would have very strong cases. So unless Wikipedia's licensing allows for that it seems very legally problematic.
Wikipedia articles are licensed under the CC BY-SA license, which (not a lawyer) appears to be compatible with what he's suggesting to do, provided there is a link to the original page, a clear indication that the original has been modified, and that the content is re-licensed under the same conditions as the original.
The license wasn't made with generative AI in mind, but the idea of basing a competing encyclopedia on modified versions of Wikipedia articles isn't inherently against the spirit of the project nor the letter of the license, provided it's done in the right way. In fact, Wikimedia Enterprise offers for-pay APIs to make it easier for corporations to re-use their content, but the for pay bit is explicitly not about paying for the right to reuse the content. It's for accessing specific APIs to make it easier to extract, work with and keep up with changes in Wikimedia content, as well as customer support for those APIs and services. The content itself is free, and Wikipedia explicitly tells corporate users not to ask them for permission to reproduce content, as that right is not Wikipedia's (it's covered by the license and the content belongs to the individual contributors, from a copyright standpoint Wikimedia owns almost none of its own content).
Okayy, I assumed the point of the Wikipedia clone was to provide Grok "unwoke" data feast upon. Isn't there an obvious problem with regression if he's still planning on doing that?
Filling it with (Elon controlled) AI garbage is precisely the point.
What's wrong with good old conservapedia?
Oh right, it doesn't have AI. Always best to have inaccuracies put in to your knowledge base at random, on purpose.
Oh wow it's worse than I thought.
https://www.conservapedia.com/Earth
I like that the "four angels standing on the four corners of the earth" as evidence for a flat earth is dismissed by "being a figure of speech", yet the same logic cannot be applied to the creation story in Genesis, that each "day" is a figure of speech.
Cherry picking season is over, go home Conservapedia.
The most worrying part is that I still can't tell if this is real or satire.
Poe's law was literally invented during an online creationism debate
Creationism is a very real "theory" and is believed by a lot of people, including my parents x.x
It would be funny if the people who believe this nonsense didn't have so much power.
My bigger takeaway is relief that "Conservapedia" is an actual website and not some derogatory name for Wikipedia. I was worried for a second that Wikipedia had some scandals and shifting ideologies/political slants I wasn't aware of. (Which, it almost certainly does, but the problems would be way worse than I thought if it got casually called that.)
The wikimedia foundation who owns wikipedia has some very liberal leaning practices, but AFAIK Wikipedia itself doesn't get involved with those. My only gripe with Wikipedia is how often they beg for donations like the site is about to shut down when the company generates insane amounts of revenue and the vast majority of their money goes to doing stuff other than running the site. That and nearly ALL of the actual work done on Wikipedia is by unpaid volunteers, so it feels a little greedy to solicit so much in donations when you're relying on free labor to run your site. (This is almost entirely because I'm jaded from being an unpaid volunteer reddit mod for a decade and watching them turn into a hundred billion dollar company off the work I did for them while simultaneously ruining the website and spitting in the mods faces time after time.)
I love Wikipedia, but whenever I get a popup to donate, it serves as a nice reminder for me to donate to the Internet Archive instead.
Same, Wikipedia is an invaluable resource supported by thousands of incredibly hard working and dedicated people, but it also isn't in any great need. There are far more deserving causes out there for you to donate to, the Internet Archive is a great example.
Does it? It's free with no subscriptions or ADs, where does the revenue come from?
[Citation needed]
It's not like they are hiding their financial reports The comments made by @OBLIVIATOR appear to be largely incorrect.
It looks like Wikimedia makes money mostly from donations and endowments, spends a large proportion of it on hosting and technology, gives out some to support it's volunteers, spends some on admin and advertising and invests the rest because they're planning/hoping to be around a long time. The only thing that seems "insane" to me about the money they make is how they managing to run one of the world's biggest websites on so little money, frankly. They were operating on less than $180m in revenue last year!
A friend of mine is a DevOps guy at wikimedia. He's not well paid by the standards of his industry, he could easily make way more working for one of the big bad tech firms but he really believes in what they're doing and I have a lot of respect for that.
I don’t know anything about how much money Wikipedia has or how it spends it, but just reading the political tea leaves tells me it would be a good idea for them to be building up a rainy-day war chest. For legal defense and technical defense alike. The Internet Archive needs this too, more urgently, as far as I can tell.
The depressing thing is that this won't stop them from harassing and trying to destroy Wikipedia.
"Fuck Wikipedia, I'm gonna make a new platform TRAINED on Wikipedia!!!"
Good luck running it on software that isn't developed by the Wikimedia Foundation (which operates Wikipedia) lol
No worries, he and his minions can use Grok to just vibe code it. And for some weird reason, variables and functions and such will reference or outright use nazi names and imagery. Just for the lolz.
Is he going to vibe-code an operating system and office suite?
Does he not know about Conservapedia?
ok elon aside, I would kinda like an AI encyclopedia, as long as it cites the sources where it got the info from and is somewhat reliable, and all articles are clearly labelled as AI generated. perplexity kinda tried but their feed is only composed of news so not very interesting
I'd be concerned about hallucinations but mostly I just don't see the value. I'm curious to where you're coming from though, what advantages do you see an AI encyclopedia offering over community curated resources like Wikipedia?
it doesn't have to be an alternative to wikipedia, it can be it's own thing. the really cool thing about wikipedia is that the community can compile all the relevant details from different sources into one article. if AI can do it, it'd be nice. because a lot of the time, the information you need is scattered around different web pages, articles, comments on reddit etc. and about hallucinations, imo perplexity has reduced the hallucinations quite a bit, I think it's reliable enough. but again, it doesn't have to be a replacement for content written by humans, just sort of something to exist alongside it. and in fact, I think it'd be pretty efficient if not everyone is generating content on it, like imagine if used the best LLMs available now (e.g. GPT 5 with the highest reasoning effort with mixture of agents). basically quality over quantity. if done right, it could actually be higher quality than some of the "educational" websites
Isn't Bing and Gemini basically that, just that the articles are generated on demand?
Surprised he isn’t calling it Xipedia, he is obsessed with X after all. Is it because “xpedia” will sound too much like “Expedia”, which is an actual company?
I guess he can afford to buy Expedia, rename it Xpedia, and turn it into an AI slop company.
Other potential names: Xwiki, WikiX, wiXi
What a dork.
Yes, but also an evil man and an enemy of humanity.
Hard agree