Newsweek: Elon Musk Takes Aim at Wikipedia

[haele]

38. However, when Wikipedia posts incorrect info -

It's typically corrected in a timely manner, and there's footnotes and attributions up the wazoo.
When AI hallucinates, it tends to perpetuate mis-information as other LLMs scrape the hallucination as fact.

I've yet to see an LLM version of AI that isn't prone to hallucination when let loose.
AI is more like a locomotive, dependent on the engineer and condition of the track it travels on.
There's still a lot of cart and horse required before AI reaches the Automobile stage where any driver can use it to get to whatever destination they need.