Elon Musk, owner of X and CEO of Tesla, has announced the release of a new AI company called xAI and its program called Grok. Photo by Bonnie Cash/UPI |
License Photo
Nov. 5 (UPI) -- A new competitor is emerging in the rapidly changing artificial intelligence virtual landscape, and it is challenging companies such as OpenAI, maker of Chat GPT, perhaps the best known AI platform to date, Inflection, Anthropic, and Google's AI platform called Bard.
xAI, the first such technology company to be released by entrepreneur Elon Musk, is unveiling Grok, which the company said is modeled on and inspired by "The Hitchhiker's Guide to the Galaxy," a 1970s British comedic science fiction program that followed the misadventures of Arthur Dent, the last person who survived the demolition of Earth by an alien race.
Clearly inspired by the fact that artificial intelligence is still often associated with futuristic science fiction, Musk also leaned on the sci-fi genre in coming up with the company name, as well.
In his 1961 science fiction novel "Stranger in a Strange Land," Robert A. Heinlein introduced the word 'grok,' which the author said had "no direct Earthling translation," but, based on the context of how the term was used in the book, critics have generally agreed that grok refers to something having very deep intuition or empathy.
The Merriam-Webster dictionary defines grok as "a transitive verb that means "to understand profoundly and intuitively."
Unlike its more conservative and buttoned-up competitors, Grok reportedly has more personality, "a bit of wit, a rebellious streak" and is wired to answer "spicy questions" that other AI apps might sidestep, according to a statement from xAI.
To prove that point, Musk promoted xAI and Grok with an unusual post leading up to the platform's release on X, the social media platform previously known as Twitter, which he also owns.
He said in the post that he asked Grok for a step-by-step cocaine recipe.
"Oh sure!" Grok responded, according to the company. "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 company said Grok also has access to information from the X database, which it claims will give xAI a leg-up on its competitors in the artificial intelligence space race.
On Sunday, Musk posted a side-by-side comparison of Grok and another AI bot answering a question. Musk claimed the competitor had less up-to-date information.
xAI, cautioned that Grok, like other Large Language Model services, "can still generate false or contradictory information," the company hedged.
Based on its limited release, Grok is not yet ready for prime time. The prototype remains in its beta testing phase and has only been in information and data recognition "training" for two months.
It is only available to a select number of users to help train and assess it during this phase and before xAI releases Grok more widely.
xAI has created a virtual waitlist where people can sign up for a chance to use the initial version of Grok before its public release. Eventually, xAI will offer an upgraded version of the program for $16 per month called X Premium+.
The company said Grok outperformed its competitors in a series of middle school math problems and Python coding tasks in a recent test. "Grok surpassed "all other models in its compute class, including ChatGPT-3.5 and Inflection-1," the company claimed, although it added Grok was outperformed by other programs with larger databases.
"In some important respects, it is the best that currently exists," Musk said of his application in an X post on Friday leading up to the Grok announcement.
Musk hired away people from OpenAI, DevelopMind and others to work for xAI, and is still hiring for several roles.
Musk has said currently available AI platforms are succumbing to pressure to be "politically correct," that xAI's mission is to create AI for people of all backgrounds and political views, and "to advance our collective understanding of the universe."
xAI calls Grok a way of testing and developing that approach to AI "in public."
Observers have warned that if not properly controlled and directed, all forms of this type of AI could be used to manipulate public opinion, spread disinformation, or amplify extremist views.
