With A 'Big Brain,' Grok 3 From xAI, Is 'The World's Smartest AI,' Said Elon Musk

The race is on, and it is getting fiercer as more powerful AIs are introduced.

Since OpenAI released ChatGPT, tech companies, large and small, compete towards supremacy, consuming raw power computation and available hardware resources to train better and smarter Large Language Models.

Elon Musk has what's called xAI, a company that creates Grok.

Months later, Grok-2 was introduced, and shows how far AIs can go when they are unhinged.

This time, xAI unveils an improved version of the chatbot, calling it the 'Grok-3.'

After teasing it days before, this marks Musk's effort to catch up with established competitors including OpenAI, Grok-3 was announced during a livestream on X.

And early tests show favorable results.

First of, according to Elon Musk himself, Grok-3 has "more than 10 times" the compute power of its predecessor.

The company went on to say that Grok-3 outperforms OpenAI’s GPT-4o, Google’s Gemini 2.0, and DeepSeek-V3 models in early testing.

And not just that because Grok-3 is also enhanced with reasoning ability.

Just like models developed by rival companies, including OpenAI’s o1, DeepSeek-R1, and Google’s Gemini Flash Thinking, Grok-3 is also able to answer more complex questions by breaking instructions down into smaller tasks and attempting to fact-check its answers before offering a solution, with the aim of providing stronger results.

Here, xAI introduces two Grok-3 reasoning modes available.

The first is called 'Think,' which will displays the AI's reasoning as it resolves requests and second, the 'Big Brain' which can solve more complex tasks by tapping into more computational power.

Another notable addition to Grok-3, is a feature called 'DeepSearch,' which the team described as an AI-powered "next generation search engine.

Musk explained that Grok-3 is a "maximally truth-seeking AI — even if that truth is sometimes at odds with what is politically correct," referring to how LLM AIs in general have been criticized for spreading election misinformation.

More or less, Grok-3 is like the U.S. answer to China's DeepSeek, which previously caught the West off guard.

Musk co-founded OpenAI in 2015, but left in 2018 due to a disagreement with CEO Sam Altman.

The billionaire then launched xAI in 2023 after the breakout success of ChatGPT, in order to show OpenAI how things should be done, at least according to him.

To make pursue this goal, he recruited talents, raised billions of dollars from investors, built a massive new data center, introduced Grok to X, a platform he also owns, and more.

While now, with xAI, Musk is capable of doing what he thinks LLMs should be developed, he still can't get along well with Altman, and still thinks OpenAI should do what he thinks.

He sued the company and Altman, alleging that the company has shifted focus from its original mission to prioritize profits.

He even attempted to buy OpenAI by leading a consortium offering approximately $97.4 billion, to which Altman declined the offer.

Regardless, Grok-3 shows immense potential and ability, and attains the top spot on Chatbot Arena, a leaderboard for AI models that is closely watched by the industry.

xAI began opening access to Grok 3 on Monday, according to Igor Babuschkin, one of xAI’s head engineers. He said the first group to get access will be X users with a Premium+ subscription. Later, xAI plans to launch a subscription tier to access advanced Grok features in its app and through the web.

Musk added that the Grok chatbot shall soon have a synthesized voice feature that sounds similar to OpenAI’s Advanced Voice Mode for ChatGPT. And meanwhile, xAI is also planning to make Grok-2 open source.