xAI, the startup led by Elon Musk that raised $6 billion in December, has a brand new AI mannequin that it claims is healthier than AI created by DeepSeek and ChatGPT-maker OpenAI.
In a live-streamed event on X on Monday that has been seen over six million occasions on the time of writing, Musk and three xAI engineers revealed Grok 3, the startup’s newest AI mannequin. They claimed Grok 3 had increased scores on math, science, and coding benchmark exams than OpenAI’s GPT-4o, DeepSeek’s V3, and Google’s Gemini AI.
Associated: Elon Musk’s xAI Is Reportedly Set to Hire Thousands of ‘AI Tutors’ With Pay Up to $65 an Hour
In addition they stated Grok 3 was a step up in sheer energy from xAI’s earlier mannequin Grok 2, launched in August. The most recent model has greater than 10 occasions the computational energy of Grok 2, better accuracy, and a much bigger capability for big datasets.
“The phrase Grok [means] to completely and profoundly perceive one thing,” Musk stated on the livestream, noting that the phrase got here from the 1961 novel “Stranger in a Unusual Land” by American creator Robert Heinlein. He added later within the livestream that “in the event you’re utilizing Grok 3, chances are you’ll discover enhancements virtually each day as a result of we’re constantly enhancing the mannequin.”
Animated 3D plot of a spacecraft launch from Earth to Mars and again. Credit score: xAI
xAI engineers demonstrated how Grok 3 may very well be used to create code for an animated 3D plot of a spacecraft launch that began on Earth, landed on Mars, and got here again to Earth.
The engineers additionally requested Grok to mix two video games, Tetris and Bejeweled, into one recreation. The outcome, which the engineers performed on the livestream, was just like Tetris with shapes inching down the display however had the principles of Bejeweled with multicolored blocks that disappeared if there have been three in a row.
Associated: Google’s CEO Praised AI Rival DeepSeek This Week for Its ‘Very Good Work.’ Here’s Why.
Musk stated that any AI may discover examples of Tetris or Bejeweled on-line and duplicate them, however Grok 3 took it one step additional.
“What’s attention-grabbing right here is it [Grok 3] achieved a artistic answer combining two video games that truly works and is an efficient recreation,” Musk famous. “We’re seeing the beginnings of creativity.”
Tetris-Bejeweled mashup recreation within the background. Credit score: xAI
The researchers stated they solely educated Grok 3’s reasoning skills on math issues and aggressive coding issues, however they noticed that Grok 3 may apply what it discovered to a wide range of use instances, together with reasoning by way of making video games.
xAI is not the one main AI startup to launch superior AI this yr. Final month, OpenAI launched the o3-mini, its most cost-effective but highly effective mannequin but, whereas DeepSeek came out with R1, a disruptive AI mannequin with cutting-edge efficiency on a lower than $6 million price range.
Grok 3 is currently available for Premium+ X subscribers paying $22 a month.
Watch the occasion, right here:
— xAI (@xai) February 18, 2025
