The following flagship AI mannequin from Elon Musk’s AI firm, xAI, could be nearing launch.
Over the weekend, a number of customers on X, together with reverse engineer Alexey Shabanov, reported getting access to Grok 3, the successor to xAI’s present mannequin, Grok 2, by way of X’s Grok chatbot app. Earlier than their entry was revoked, the customers say that they managed to get the mannequin to reply to a spread of queries, together with logical reasoning and coding-related questions.
In response to the customers, Grok 3 was in a position to efficiently reply riddles and area requests like “Generate HTML and JavaScript code for a roulette wheel on line casino.” The mannequin wasn’t good, nevertheless. Within the code for the roulette wheel, Grok 3 missed a couple of particulars — and made a programming error.
Breaking 🚨 Grok 3 understands nuanced riddle that ALL of open ai fashions have didn’t get proper. The one different mannequin to get this query proper was Deepseek R1 🤯
Grok 3 is thrashing o1? Has @xai caught up this quick ? pic.twitter.com/WLSpM00qHK
— Chris (@chatgpt21) January 26, 2025
Shabanov mentioned that he obtained Grok 3 to disclose its system immediate, which is the set of directions that outline how the mannequin ought to behave. One attention-grabbing element is that the immediate seems to explicitly state Donald Trump is the forty seventh president of the U.S., which may very well be a hardcoded repair for Grok’s political hallucinations.
Grok 3, which has been in improvement for a number of months, was optimistically slated for launch final yr, in keeping with Musk — however missed that deadline. Early this month, Musk mentioned that Grok 3 had accomplished pre-training, a vital stage in a mannequin’s improvement cycle, and that he anticipated Grok 3 to reach in January or early February.
xAI has been utilizing its huge information heart in Memphis — a knowledge heart containing round 100,000 GPUs — to coach Grok 3. In a put up on X, Musk claimed that Grok 3 was skilled with “10x” extra compute than Grok 2.
Grok, xAI’s reply to fashions like OpenAI’s GPT-4o and Google’s Gemini, can analyze photos and reply to questions, and powers a lot of options on X. Grok just lately gained standalone apps, and it might quickly get a “voice mode” that reads the mannequin’s responses aloud.
When Musk introduced Grok roughly two years in the past, he pitched the chatbot as edgy, unfiltered, and anti-“woke” — basically, prepared to reply controversial questions different AI methods gained’t. He delivered on a few of that promise. Advised to be vulgar, for instance, Grok will fortunately oblige, spewing profanities and colourful language you wouldn’t hear from ChatGPT.
However Grok because it exists in the present day hedges on political topics and gained’t cross sure boundaries. The truth is, one examine discovered that Grok leans to the political left on matters like transgender rights, range applications, and inequality.
Musk has blamed the habits on Grok’s coaching information — public webpages — and pledged to “shift Grok nearer to politically impartial.” Proof of an “Unhinged Mode” for Grok just lately emerged, which is able to present responses “supposed to be objectionable, inappropriate, and offensive,” in keeping with an FAQ web page on xAI’s web site.
Musk has additionally mentioned that Grok 3’s coaching information incorporates filings from courtroom circumstances, bettering its means to know authorized topics.
