Elon Musk’s startup xAI has unveiled its first product, a large language model with chatbot capabilities named Grok, currently available via an early access waitlist with plans to go wide to Premium+ subscribers to the X social platform (formerly Twitter) following beta tests. The company says Grok has “access to search tools and real-time information” and is extremely up-to-date, but “as with all the LLMs trained on next-token prediction, our model can still generate false or contradictory information.” The chatbot is distinguished by sarcasm and wit, “so please don’t use it if you hate humor,” xAI warns.

Musk, in social-media posts on X over the weekend, “said Grok will be designed for tasks including information retrieval and coding assistance, part of an effort to create AI tools ‘that assist humanity in its quest for understanding and knowledge,’” writes The Wall Street Journal.

Grok is modeled after writer Douglas Adams’ “The Hitchhiker’s Guide to the Galaxy,” “so intended to answer almost anything and, far harder, even suggest what questions to ask,” according to an xAI news announcement.

Along those lines, the Grok chatbot “has a rebellious streak” and often seeks to get a laugh, a skill Musk demonstrated by sharing Grok jokes about FTX founder Sam Bankman-Fried, according to WSJ, which writes that Musk has “denounced OpenAI as being politically correct, among other things” and is offering up Grok as “an alternative AI option to Google and to Microsoft, which is a major investor in OpenAI.”

Musk posted on X that Grok will “answer spicy questions that are rejected by most other AI systems.” Business Insider shares xAI staffer Toby Pohlen’s demo of “a regular setting and ‘a fun mode.’”

Musk stressed that Grok is “still a very early beta product — the best we could do with two months of training — so expect it to improve rapidly with each passing week.” Grok is powered by the Grok-1 foundation model.

Since April, when word began leaking out that Musk was launching his own artificial intelligence venture, xAI “trained a prototype LLM (Grok-0) with 33 billion parameters. This early model approaches LLaMA 2 (70B) capabilities on standard LM benchmarks but uses only half of its training resources,” the company explains in its announcement.

The last two months have seen “significant improvements in reasoning and coding capabilities leading up to Grok-1, a state-of-the-art language model that is significantly more powerful, achieving 63.2 percent on the HumanEval coding task and 73 percent on MMLU.”

VentureBeat — noting that Grok is slang for “to understand” — parses the benchmark tests cited in the Grok announcement, explaining xAI claims that Grok “surpass[es] all other models in its compute class, including ChatGPT-3.5 and Inflection-1,” and says “it is only surpassed by models that were trained with a significantly larger amount of training data and compute resources like GPT-4.”