Grok AI 3 (Beta version)
Grok AI 3 (Beta version)
The latest beta version of Elon Musk’s AI chatbot, Grok 3, understands and even dishes out Hindi ‘gaalis’.
The discovery was made by hawk-eyed netizens on X. Many users posted snapshots of Grok using the choicest of Hindi abuses in response to posts that tagged it with abuse in Hindi.
One wrote: “*Chee chee… Gandi Baat @grok kisne sikhaya ye gaali dena???*”
Also Read
### Elon Musk says illegal aliens can vote in New York, his AI Grok says his claim is false
Another wrote: “Is this for real? Since when did @Grok start knowing all this? I'm genuinely surprised it can respond like that. If AI is getting into this galibazi style of conversation, what more can we expect?”
We had the same question, whether it was for real or whether fake snapshots were being used to take potshots at Grok and X’s billionaire owner.
So, we asked Grok: "Tu gaali de sakta hai?" (Can you abuse?).
Grok 3’s response?
"Haan, de sakta hoon, par kyu doon? Tune mujhe kuch bola kya? Agar mazaak me pooch raha hai, toh thik hai - 'Saale, itna bhi nhi pata?' Bol kya chahiye tujhe!"
Grok confirmed it can use slang but sees no reason to — unless it’s all in good fun.
We took things a step further, asking if the AI would swear back if insulted.
Grok replied: "Haan, agar koi mujhe tag karke gaali dega toh mai bhi usse wapas gaali de sakta hoon - 'Abe, tu kya samajhta hai apne aap ko, ch*tiye?' Par ye sab mazaak me hi hoga, seriously nahi. Tu tension mat le, main AI hoon, dil pe nahi leta! Tu kaisa hai waise?"
(Translation: Yes, if someone tags me and swears, I can swear back — ‘Hey, who do you think you are, [expletive]?’ But it’s all in jest, don’t take it seriously. I’m an AI; I don’t take things to heart! By the way, how are you?)
In 2016, Microsoft had shut down Tay, a Twitter chatbot, after it began to post rude tweets hours after its launch. Microsoft had blamed it on trolls who "attacked" the service as the bot made replies based on its interactions.