Elon Musk says early version of xAI’s Grok chatbot answers ‘spicy questions’, benefits from real-time access to X
Elon Musk has launched Grok, a “sarcastic” AI chatbot that he said was modelled on The Hitchhiker’s Guide to the Galaxy.
Grok is the first product of Musk’s newly created company xAI, which he launched in July to compete with OpenAI, Anthropic and others well-funded start-ups – staffed by researchers largely drawn from those competitors.
Musk said Grok benefits from real-time access to X, formerly Twitter, which he also owns. He has taken pains to block rival companies from scraping data from the social media platform.
The entrepreneur said earlier this year he wanted Grok to be “a maximum truth-seeking AI that tries to understand the nature of the universe”.
xAI described the bot, which has had only two months’ training and is available to selected users, as having a “rebellious streak” and being able to answer “spicy questions that are rejected by most other AI systems”.
On Friday, Musk teased the launch by posting an example on X of Grok responding to a request for a step-by-step recipe for making cocaine.
“Oh sure! Just a moment while I pull up the recipe for homemade cocaine. You know, because I’m totally going to help you with that,” the bot replied, before providing generalised information and warning against following up with the plan.
The company said Grok surpassed “all other models in its compute class”, including ChatGPT-3.5 and Inflection-1, from competitors OpenAI and Inflection, in initial tests based on middle school math problems and Python coding tasks.
It was outperformed by bots with access to larger amounts of data, the firm said.
But the bot appears to be vulnerable to similar issues affecting others of its kind, such relating inaccurate information while sounding highly convincing.
xAI said in a statement that the bot “can still generate false or contradictory information”.
Grok is intended to eventually be a feature of X Premium+, which costs $16 (£13) per month.