On October 27th, 2025, Musk launched Grokipedia. Musk’s love for black shines through the interface of Grokipedia, but it might be because I’m looking at it at the wee hours when it’s still dark outside. It could be imitating the light conditions where I am. After all, it’s powered by AI.
It’s got an ever-growing number of articles in its database, which, as of October 31st, was:
Anyway, let’s come to the point.
The Genesis of Grokipedia
Musk has had a love-hate relationship with Wikipedia. According to the Grokipedia page on Wikipedia, Musk has said that he’d give a billion dollars to Wikipedia, if they were willing to name it “Dickipedia.” Later, in 2024, he called it “Wokepedia” (Woke is a term that was once used to describe broadmindedness. However, in recent years, it’s been increasingly used to refer to being “overly progressive.)
While Grok, by Musk’s own admission, has “learned” a lot from Wikipedia, in September 2025, the idea of Grokipedia was born, and now, end October, we already have it with us. (Time from idea to execution: 1 month.)
As it is expected of any new AI product/service, Grokipedia is facing its share of criticism that falls into the usual three categories: 1. Biases, 2. Hallucinations, and 3. Copying.
We’ll have to wait to see how Grokipedia fares against Wikipedia. The wait isn’t going to be long, I suppose – because anything AI-related just happens at lightening speed.
When I want to find information, Wikipedia often serves as my primary node. It gives me an excellent overview of the topic in question. It also lets my curiosity guide my learning.
Grokipedia, on the other hand, is fairly structured and controls the flow of your learning quite like the ancient web pages did. Linearly.
My Experience with Grokipedia
Here’s an example. Since I had been watching a Netflix documentary on Aileen Carol Wuornos, a woman serial killer, I was curious about other women serial killers. I typed the term “women serial killers” in Grokipedia’s search bar, and it assumes that I am seeking a list, so it provides me a list of links (French WSKs, German WSKs, Spanish WSKs, and so on.)
When I click on any of the links, I get a table with the names and their quick descriptions along with their number of kills, active years, and current status. And, of course, the reference link, embedded as a non-distracting number right at the end of the sentence. Perfect citation.
Also, each information page ends with “References” quite like in the case of Wikipedia.
My Grokipedic Pain Points
So what’s not sitting right? (Not Li’l Bit, who appears to be comfortable perched on his beanbag.)
Everything’s good if you are making a skim-the-surface report and need some basic data (that must be verified anyway in both cases). However, it doesn’t let you learn the way you, as a human, is supposed to learn. Make the basic connections, go in a direction of your interest, dig; then return, read again, then take off in another direction. Later, synthesize all that stuff and form a perspective.
But that’s just one part of it. There’s more.
- The white on black. (Wikipedia is black on white.)
- The lack of graphics. (Wikipedia has a strong visual dimension.)
- The lack of connecting Grokipedia pages. (Wikipedia is a dense web of interconnecting pages)
- The assumptions that Grok makes (I didn’t ask for a list of serial killers, categorized by nationality). A similar search on Wikipedia (Female Serial Killers) takes me to the portion of their “serial killers” page, where I learn about female serial killers, their rarity as serial killers (only about 16% of them are women), and a lot more.
- An eerie feeling of clouds gathering on the horizon. (This one’s totally on me.)
End Note:
I’m staying with Wikipedia, but I’ll definitely keep a tab on Grokipedia. I’d like to see how it evolves and how it becomes more useful than Wikipedia.
Image Credits:
Screenshots: Grokipedia
Li’l Bit: Dad: ChatGPT/Uncle: Gemini
Leave a Reply