Wikipedia founder Jimmy Wales is brushing off Elon Musk's AI-powered Grokipedia challenge, arguing that large language models are fundamentally unfit to create reliable encyclopedic content. Speaking at CNBC's Technology Executive Council Summit, Wales dismissed Musk's claims of Wikipedia bias while highlighting critical flaws in AI-generated information that could undermine public trust in digital knowledge sources.
Elon Musk's latest venture into content creation hit immediate skepticism from an unexpected source - the very platform he's trying to replace. Wikipedia founder Jimmy Wales didn't mince words about Grokipedia during Tuesday's CNBC Technology Executive Council Summit in New York, delivering a pointed critique that cuts to the heart of AI's reliability crisis.
"I'm not optimistic he will create anything very useful right now," Wales told the audience, dismissing Musk's AI-powered Wikipedia alternative before even examining its output. The comment comes as Grokipedia struggles through what Wales called a "rocky start" in its public debut.
Wales' skepticism isn't personal - it's technical. The Wikipedia founder has spent years watching large language models fail at the very task Grokipedia claims to master. "The LLMs he is using to write it are going to make massive errors," Wales explained. "We know ChatGPT and all the other LLMs are not good enough to write wiki entries."
The economic disparity between the two approaches is staggering. Wales estimates Wikipedia's annual technology costs at just $175 million, while Wall Street projects AI hyperscalers will spend $550 billion next year alone. Yet despite this massive investment gap, Wales argues the AI approach fundamentally misunderstands the problem it's trying to solve.
To prove his point, Wales shared real-world examples of LLM failures that expose the technology's core weaknesses. When he tests new chatbots by asking about his wife - "not famous but known" from British politics - the results are consistently "plausible but wrong." The pattern reveals how LLMs create convincing fiction when pressed for specific details.
Even more damaging was Wales' account of a German Wikipedia community member who discovered systematic citation errors. The volunteer wrote a program to verify ISBN numbers in book references and traced multiple fake citations back to a single contributor. That person eventually confessed to using ChatGPT for research - and the AI "just very happily makes up books for you," Wales noted.
The bias accusations that motivated Musk's Grokipedia launch drew particular fire from Wales. "He is mistaken about that," Wales said of Musk's "woke bias" claims. "His complaints about Wiki are that we focus on mainstream sources and I am completely unapologetic about that. We don't treat random crackpots the same as The New England Journal of Medicine and that doesn't make us woke."
Wales couldn't resist a jab at Grokipedia's apparent self-promotion: "I haven't had the time to really look at Grokipedia, and it will be interesting to see, but apparently it has a lot of praise about the genius of Elon Musk in it. So I'm sure that's completely neutral."
The competitive threat extends beyond just Musk's project. Wales expects the misinformation challenge to intensify as LLMs become better at creating fake websites with plausible content. While he believes Wikipedia's experienced community of editors can spot these deceptions, "it will fool a lot of people and that is a problem."
Interestingly, Wales isn't entirely opposed to AI assistance. He's been exploring limited applications where AI might help Wikipedia editors find additional information in existing sources - a use case he describes as "kind of okay." The key difference is human oversight and verification, something automated wiki generation lacks.
Wales frames Wikipedia's role as critical infrastructure, acknowledging the responsibility that comes with that position. "We are really happy Wiki is now part of the infrastructure of the world, which is a pretty heavy burden on us," he said. "So when people say we've gotten biased, we need to take that seriously."
But he couldn't resist one final dig at the competition: "We talk about errors that ChatGPT makes. Just imagine an AI solely trained on Twitter. That would be a mad, angry AI trained on nonsense."
The Grokipedia challenge highlights a broader tension in the information ecosystem. While AI companies pour billions into language models, Wikipedia has built the world's largest encyclopedia through human collaboration and rigorous sourcing standards - at a fraction of the cost.
Wales' dismissal of Grokipedia reflects deeper concerns about AI's readiness for knowledge curation. While Musk's team pours resources into automated content generation, Wikipedia's human-driven model continues proving its worth through accuracy and community oversight. The real question isn't whether AI can compete with Wikipedia, but whether the public will recognize the difference between generated content and verified knowledge. As misinformation tools become more sophisticated, Wales' 25-year-old collaborative model may be more valuable than ever.
