Elon Musk has entered the knowledge arena, launching Grokipedia, a new online encyclopedia powered by his xAI models. As its name suggests, it is a direct competitor to the massively popular Wikipedia. In fact, Musk claims Grokipedia will “purge out the propaganda” found on Wikipedia and exceed its rival “by several orders of magnitude in breadth, depth, and accuracy.”

However, the man who co-founded Wikipedia, Jimmy Wales, disagrees with that statement. Speaking at a recent CNBC summit, Wales dismissed Grokipedia’s threat, arguing that the underlying technology—Large Language Models (LLMs)—simply isn’t capable of producing a reliable encyclopedia.

Wikipedia’s Jimmy Wales is not worried about Elon Musk’s AI-powered Grokipedia

Wales’ skepticism stems from Grokipedia being powered by AI. He is aware of the flaws of LLMs like Grok and ChatGPT, which have a tendency to “make massive errors” and fabricate sources—namely “AI hallucinations.” Wales shared an example where a Wikipedia community member found that a source had used an LLM to generate book citations. This allegedly resulted in the chatbot “very happily mak[ing] up books.”

To test LLMs, Wales often queries obscure topics, such as the identity of his wife (a “not famous but known” figure in British politics). He claims that this search consistently results in a plausible but ultimately wrong answer.

“We know ChatGPT and all the other LLMs are not good enough to write wiki entries,” Wales stated. He emphasizes that the reliance on human verification and community oversight that Wikipedia has developed over decades.

The question of bias and sources

Musk and his allies frequently criticize Wikipedia for alleged “woke bias” and for excluding conservative media as sources. Grokipedia’s own entry on Wikipedia accuses the site of having a “systemic ideological bias.”

Wales strongly defended Wikipedia’s editorial policy. He said the organization’s priority is sourcing neutrality. “We don’t treat random crackpots the same as The New England Journal of Medicine, and that doesn’t make us woke,” Wales said. He also added that responding to criticism by “doubling down on being neutral and being really careful about sources” is essential.

The Wikimedia Foundation still acknowledges that AI poses challenges—notably the decline in human visits and the increased scraping of Wikipedia data to train AI models. However, he considers Wikipedia’s approach to be better and more reliable. “This human-created knowledge is what AI companies rely on to generate content; even Grokipedia needs Wikipedia to exist,” notes the Foundation.

Ultimately, Wales views the biggest threat not as a direct AI replacement. Instead, he sees it as the ability of LLMs to generate plausible fake websites, which could fool the general public.

The post Elon Musk's Grokipedia Launches, but Wikipedia Isn't Scared appeared first on Android Headlines.