Behold 'Grok,' Elon Musk's AI chatbot with a 'rebellious' streak

Power Play: Cybersecurity chiefs on misuse of AI Canadian and U.S. cybersecurity chiefs on what both nations are doing to combat bad actors from misusing artificial intelligence.

London — Brace yourselves. Jokes incoming. Elon Musk has announced the introduction of “Grok,” an artificially intelligent chatbot, for some users of X, which the billionaire suggests has a sarcastic sense of humour similar to his own.

Musk, who has owned X — formerly known as Twitter — for a year, announced late Saturday that Grok is being trained by having “real-time access” to information from the platform.

Musk’s AI startup, xAI, which developed the bot, said in a blogpost that it took inspiration from The Hitchhiker’s Guide to the Galaxy, a comedy sci-fi novel by British author Douglas Adams.

Grok “loves sarcasm,” Musk posted on X. “I have no idea who could have guided it this way.”

“Grok is designed to answer questions with a bit of wit and has a rebellious streak, so please don’t use it if you hate humour!” xAI said in its blogpost.

The bot is still in an early stage of testing, Musk said, but will soon be made available to subscribers of X’s Premium+ service, who, in the United States, pay $16 a month for features such as the once-coveted blue checkmark.

Grok is intended to rival ChatGPT, an AI chatbot released to the public by its maker OpenAI last November to much fanfare. Musk is a co-founder of OpenAI, but stepped down as chairman five years ago partly over disagreements about the company’s direction.

In its blogpost, xAI said the large language model (LLM) — the type of system used to train AI applications — underpinning Grok-1 had outstripped the capabilities of GPT-3.5, the LLM that trains the publicly available version of ChatGPT, on several measures, including solving some high school math problems.

“It is only surpassed by models that were trained with a significantly larger amount of training data and compute resources like GPT-4,” the company said.

Musk has previously expressed concerns about the speed of development of AI, and the technology’s potential to fundamentally change human society for the worse.

xAI said it expects Grok to rapidly improve as X’s users feed it with more information, which may fuel unease among those critical of Musk’s handling of the platform since he bought the company last October.

The world’s richest person has called himself a “free speech absolutist” and said that, under his leadership, the company would be “very reluctant” to delete content and permanently ban some users.

British far-right figure Tommy Robinson, who has a history of posting Islamophobic content on X, thanked Musk Sunday for reinstating his account, five years after being banned.