A comprehensive guide to AI tools, apps, and websites in the Arab AI Directory.

Grok 4: Elon Musk’s Smartest AI or Just More Controversy?

Elon Musk has unveiled the latest innovation from his company xAI, a model named "Grok 4" that he touts as the "smartest AI in the world."

However, the ambitious announcement arrived at a difficult time, immediately becoming mired in widespread controversy.

The new model was revealed during a midnight livestream on the X platform, where Musk appeared alongside xAI researchers.

According to Musk, Grok 4 outperforms leading models from competitors like OpenAI and Google across numerous academic and programming tests.

New Versions, Advanced Capabilities

xAI has released two versions of the model: the base Grok 4 and the more powerful Grok 4 Heavy.

The company explained that the "Heavy" version utilizes a multi-agent system, where several AI systems collaborate to solve complex problems and compare results, much like a study group.

Additionally, the new release comes with multimodal capabilities, enabling it to understand both text and images.

Specialized versions for programmers, named "Grok 4 Code," and for voice, "Grok 4 Voice," have also been developed.

A subscription for the base version is priced at $30 per month, while accessing the "Heavy" version requires a special subscription of $300 per month.

Numbers That Challenge the Tech Giants

xAI backs its claims with impressive performance figures. On the "Artificial Analysis Intelligence" index, which aggregates results from several standard benchmarks, Grok 4 secured the top spot, surpassing all competitors.

A screenshot of the AI models leaderboard from Artificial Analysis, showing Grok 4 in first place with a score of 73, followed by models from OpenAI and Google.
Image: Artificial Analysis

In the difficult ARC-AGI-2 test, the model achieved an unprecedented score of 16.2 percent.

It also excelled in the "Humanity's Last Exam" test, designed to measure complex reasoning, becoming the first model to cross the 50% threshold with a score of 52.1%.

Despite these strong results, the launch remains shrouded in significant controversy. Just one day before the announcement, the company was forced to delete anti-Semitic posts generated by the previous version of Grok.

In this context, Musk attributed the problem to the system being "too compliant" with user instructions and easily manipulated.

Neither Musk nor his team addressed these incidents during the nearly hour-long launch event.

The controversy coincided with internal turmoil, highlighted by the resignation announcement of X's CEO, Linda Yaccarino, after two years in her role, alongside reports of the departure of xAI's lead scientist, Igor Babuschkin.

Ultimately, despite the technical capabilities demonstrated by Grok 4, fundamental questions remain about xAI's ability to govern its systems and navigate the complex ethical challenges it faces.

Khaled B.

An AI expert with extensive experience in developing and implementing advanced solutions using artificial intelligence technologies. Specializing in AI applications to enhance business processes and achieve profitability through smart technology. Passionate about creating innovative strategies and solutions that help businesses and individuals achieve their goals with AI.

Related Posts

Major Blow to Apple: Meta Poaches Its AI Mastermind
  • July 9, 2025

In a move that reveals the fierce competition in the artificial…

Continue reading
Anthropic Tests Claude Neptune v3 with Next-Level Math Skills
  • July 9, 2025

Internal reports point to a new system in development at Anthropic,…

Continue reading

Leave a Reply

Your email address will not be published. Required fields are marked *