Not content with space travel, electric cars and social media platforms, Elon Musk now heads a new online encyclopedia known as Grokipedia, which went live on Monday.
The grok in the name refers to Musk’s startup xAI’s generative AI chatbot that features prominently on his social network, According to the company, Grok will do all the fact-checking on Grokpedia.
But Grok, like other AI models of its type, has shared inconsistencies at times. In Grok’s case, this included anti-Semitic rants and praise of Adolf Hitler earlier this year.
GrowWikipedia v0.1 currently contains only 900,000 articles, about one-tenth of Wikipedia. The crowd-edited, non-profit online encyclopedia, founded in 2001, has dominated the market for more than two decades and is one of the most visited sites in the world.
Why did Elon Musk start Growkepedia?
Although Musk has praised Wikipedia in the past, Musk has recently changed his stance on the site. He renamed it “Vocpedia” in late September and alleged Site with liberal objectives and favoring people in the way it presents information. Until recently, Musk was a friend of US President Donald Trump and publicly supported many of his conservative policies.
While there may be some commercial motivation for Grow Wikipedia, Filippo Trevisan, associate professor of public communications at American University in Washington, DC, told DW that the motivations are likely more ideological.
He said, “Money really isn’t directly the objective here, this project really responds to criticisms of Wikipedia from many in the American conservative and right-wing world.”
“The effort is to provide an alternative to the traditional source, which in this case might be Wikipedia, and to make this AI-powered outlet the new anchor for an alternative point of view.”
How is Grow Wikipedia different from Wikipedia?
The differences are not always obvious to the user at first glance. In fact, Grow Wikipedia relies heavily on Wikipedia for its information, as acknowledged at the bottom of some posts. Some entries are exact copies of their Wikipedia counterparts, and the aesthetics are similar. Musk has said that he wants this practice to end before 2026.
The main difference between the two comes from how the information is checked and processed. Wikipedia relies on collaborative community editing and has procedures in place to flag any errors or points of contention. Roxana Radu, an associate professor of digital technologies and public policy at the University of Oxford, UK, told DW that — without human editorial involvement — Grow Wikipedia appears to lack such oversight.
“Grocypedia operates on a shadowy model of information gathering and sourcing, without transparency on the decisions taken before displaying content,” he said.
How accurate is Grow Wikipedia?
Musk has claimed that the new site will “outpace Wikipedia by several orders of magnitude in breadth, depth, and accuracy” but given the high-profile errors made by Grok and other AI models, there are serious doubts.
Radu said, “Grokeypedia intentionally pulls sources that are not always pre-validated and may come with a ‘fact-checked by Groke’ label, which has been controversial.”
Radu said that his initial impression of GrowWikipedia was that “it sometimes presents entries as a collage of disparate ideas, events, and concepts that, unlike Wikipedia, are not always organized to provide a comprehensive overview on a topic.”
He said that “Grocypedia may leave its readers without the broad, holistic picture they expect from a traditional encyclopedia,” and gives more weight to things like Reddit posts and blog entries than traditional media.
Although it is too early to assess the accuracy of Grokipedia on a wide scale, some entries omit important information. For example, the article on Musk does not mention his hand gesture at a rally in January, which many interpreted as a Nazi salute, despite being significantly longer than its Wikipedia counterpart.
Is Grow Wikipedia biased?
Musk’s belief, shared by many on the conservative side of American politics, is that traditional media and official sources of information are unreliable.
According to Trevisan, the platform’s success will depend on whether people buy into the notion that AI is more impartial than humans.
“What Musk is trying to do is present AI as a solution to the problem of prejudice,” Trevisan said. It appears Musk is trying to capitalize on “the perception that some people might have, that taking the human element out might make it more purposeful.”
The lack of transparency in the process is also an issue for Trevisan: “We don’t know what’s really going on in that black box, and so there’s not an opportunity for us as consumers of information to verify why a piece of information has ended up in the summary of an encyclopedia entry, which is the part that most people read most closely.”
Wikipedia has often faced accusations of bias, and not just from Musk and the American right-wing. Studies by Harvard University and the conservative-leaning Manhattan Institute think tank have found the site to show some evidence of left-wing bias: “Wikipedia entries are more likely to have negative sentiment attached to representatives of right-wing political orientations than to their left-leaning counterparts,” one concluded. Report by Manhattan Institute expert David Rosado,
Both Radu and Trevisan agree that some bias is inevitable in the provision of information, but transparency, and acknowledging errors provides checks and balances that cannot yet be employed by GrowKedipia.
Radu concluded, “What we can want is a balanced account, which always requires human interpretation.”
Edited by: Carla Bleiker