Through his official X account on Friday, Elon Musk announced a new feature for xAI’s latest artificial intelligence model, called “Ask Grok,” designed to reply directly to posts on the social media platform. The feature allows users to tag Grok in any post or reply, prompting the AI to generate responses based on the content.
Musk publicized the update after Aditya Paliwal, an artificial intelligence software engineer at xAI, tested the chatbot by posting an image of an anime character from the 1989 Japanese anime series Dragon Ball, asking Grok to identify it. The AI recognized the character as Vegeta, also known as the Prince of Saiyans in the franchise.
@grok who is this character?
— Aditya Paliwal (@VastoLorde95) March 7, 2025
The Tesla CEO quoted the post saying: “Ask @grok is now live. Just reply to a post with @grok!”
Ask Grok is live, but somewhat unresponsive
In response to Paliwal’s post, Grok named Vegeta and gave more information about the shirt he was wearing by saying:
“He’s depicted wearing a pink shirt, a humorous reference to a scene where Bulma, his wife, jokingly dresses him in casual attire. Dragon Ball, created by Akira Toriyama, follows the adventures of Goku, his friends, and Vegeta, who was a rival turned ally.”
Another X user asked the chatbot to provide a summary of Vegeta’s story. Grok replied with a one-paragraph-long character biography, explaining Vegeta’s origins as a Saiyan prince born in Age 732 to King Vegeta.
It also talked about his early service under the villain Frieza, his eventual defection to Earth, and his character development from a ruthless warrior to a family man, married to Bulma and father to Trunks and Bulla.
Other users tried to engage the AI model with questions of their own, but it was mostly unresponsive, which could have been caused by too many attempts from netizens trying to test out the feature.
“Grok doesn’t seem to work correctly after I’ve thumbed through the comments here,” one netizen complained.
Several conversational chatbots are currently active on X, including another US-based model, Perplexity AI. Just like Grok, users can tag “@AskPerplexity” in their posts to receive answers or check facts on the social platform.
Still, some claim Perplexity occasionally struggles to interpret viral memes or occurrences in the world that requires it to be familiar with what has happened online.
Ukrainian developer tests Grok 3 to create language learning app
In other news, a Ukrainian observability engineer named Yevgeny recently used Grok 3 to assist in creating a simple English learning application for his younger brother.
In a translated comment shared by Ukrainian tech publication dev.ua, Yevgeny said that he is not a professional programmer, but wanted to develop something useful for his brother, who is learning English. Using JavaScript, he built an English Words Trainer app with guidance from Grok 3.
According to Yevgeny, the AI felt “like a smart friend who really wants to help,” offering assistance throughout the development process.
Grok-3 appears to mimic Elon Musk’s conversational and sometimes sarcastic tone when given any prompt, much more different than OpenAI’s ChatGPT, which focuses more on neutrality.
During a three-hour conversation with podcaster Joe Rogan last week, Elon Musk had Rogan ask the chatbot whether “all the gold was in Fort Knox,” to which Grok responded: “Are you a f******* conspiracy theorist?” making both Rogan and his guest crack up.
Cryptopolitan Academy: Want to grow your money in 2025? Learn how to do it with DeFi in our upcoming webclass. Save Your Spot
