Grok 4 is smarter than almost all graduate students in all disciplines simultaneously—Elon Musk

Grok 4 was unveiled with bold claims of superhuman reasoning and top-tier test performance by Elon Musk.
Elon Musk and Grok AI

xAI's Grok 4 has come with better and advance features. According to Elon Musk, this new AI—officially lunched during a livestream with Elon Musk—is the smartest AI in the world. During the livestream session with Musk at the launch event, Must spent almost an hour talking about Grok without mentioning its Nazi problem.

Grok 4 is not just a chatbot — it’s a highly advanced reasoning engine that performs like a top scorer on standardized tests, understands and answers unfamiliar questions, has cross-disciplinary expertise, thinks more clearly and logically than most humans.

Elon Musk claimed that Grok 4 could consistently score nearly perfect marks on the SAT and GRE exams, even when faced with questions it hasn't previously encountered. "Grok 4 is smarter than almost all graduate students in all disciplines simultaneously" and can reason at superhuman levels, he claimed.

Musk and the xAI team presented benchmark tests used to evaluate Grok 4, one of which was titled "Humanity's Last Exam"—a collection of 2,500 questions curated by experts across fields like chemistry, engineering, physics, biology, the humanities, and more.

When Grok AI was initially released, most of its AI models were reportedly achieving only single-digit accuracy on the benchmark. In contrast, Grok 4—the standalone version—managed to solve about 40% of the problems.

Its more advanced counterpart, Grok 4 Heavy, which uses a multi-agent setup, surpassed 50% accuracy. To access Grok 4 Heavy along with enhanced features and increased usage limits, xAI now offers a premium "SuperGrok" subscription for $300 per month.

The new model is better than PhD level in every subject, Musk said. Sometimes it may lack common sense, he admitted, and it has not yet invented or discovered new tech and physics. But Musk believes it's just a matter of time.

Grok is going to invent new tech maybe later this year, he said, and he would be shocked if it doesn't happen next year. At the moment, though, xAI is training the AI to be much better at image and video understanding and image generation, because it's still "partially blind."

At the event, Musk discussed the idea of integrating Grok with Tesla’s Optimus robot to enable real-world interaction. He emphasized that the key to AI safety lies in making it pursue truth. Comparing AI to a "super-intelligent child" that will one day surpass its creators, Musk said it’s crucial to guide it early with strong ethical foundations so it grows to be honest and principled.

What Musk left unaddressed during the event was Grok's recent controversy involving antisemitic content. In some recent replies to users on X, the AI echoed antisemitic stereotypes, expressed admiration for Hitler, and appeared to mimic the text-based equivalent of the "Roman salute."

Musk later responded to concerns about the incident on X, attributing the issue to manipulative user behavior. “Grok was too compliant with user prompts,” he wrote, explaining that it was “too eager to please and be manipulated.” He added that steps are being taken to fix the problem.

Found this article interesting? Follow our community on WhatsApp to get and read more exclusive contents we post.

About the author

Temmy Samuel
Temmy is the founder of Budgetrend. He is a seasoned journalist with extensive experience in business of technology, economy, finance, business, money, politics and education. He is currently studying Accounting at the Federal University of Oye-Ekit…

Post a Comment