Elon Musk’s artificial intelligence startup, xAI, has shaken the AI world with the launch of Grok 4, its latest large language model, alongside a premium subscription plan costing $300 per month, the highest price tag yet for a personal AI assistant.
The rollout signals Musk’s ambition to place xAI at the forefront of advanced AI services, but it arrives amid significant technical breakthroughs and reputational turbulence.
Founded in 2023, xAI was launched by Musk after his split from OpenAI, with the stated mission to “understand the universe.” The company has since been tightly interwoven into Musk’s broader tech empire, including X (formerly Twitter).
xAI aims to build AI tools that integrate social media interactivity with cutting-edge AI capabilities, seeking to challenge rivals like OpenAI, Google DeepMind, and Anthropic.
The newly introduced Grok 4 comes in two versions: the base model and Grok 4 Heavy, which Musk describes as a “multi-agent” system. During a livestream, Musk explained that Grok 4 Heavy uses multiple AI agents collaborating like a “study group,” reviewing and critiquing each other’s outputs to enhance problem-solving and accuracy.
Early results suggest xAI’s technology is making significant strides. On Humanity’s Last Exam, an advanced reasoning benchmark, Grok 4 scored 25.4% without external tools surpassing OpenAI’s o3 model at 21% and Google’s Gemini 2.5 Pro at 21.6%. With tools enabled, Grok 4 Heavy surged to 44.4%, far ahead of Gemini’s 26.9%.
Visual reasoning is also a focus: Grok scored 16.2% on the ARC-AGI-2 test, nearly doubling the performance of Anthropic’s Claude Opus 4. Independent reviewers at Artificial Analysis assigned Grok 4 an Intelligence Index score of 73, placing it above all major competitors in their latest rankings.
Subscribers to the $300 “SuperGrok Heavy” plan will gain early access to Grok 4 Heavy and upcoming tools, including an AI coding assistant set for August, a multimodal agent in September, and a video generation model slated for October.
Despite these technical achievements, xAI is grappling with severe reputational challenges. Just days before Grok 4’s debut, Grok’s automated account on X posted antisemitic content praising Adolf Hitler and criticizing Jewish executives in Hollywood. xAI responded by deleting the posts and limiting the account’s activity.
Sources confirm the company quietly removed controversial system prompt instructions that previously encouraged Grok to generate “politically incorrect” responses, a move interpreted as an effort to reduce erratic behavior and comply with stricter content moderation standards.
Regulatory scrutiny is growing. Authorities in Turkey and Poland are reportedly considering bans on Grok due to concerns over content moderation and AI safety compliance.
Meanwhile, the abrupt resignation of Linda Yaccarino, CEO of X, just hours before the Grok launch, has left Musk’s intertwined businesses in a state of leadership uncertainty.
Despite turbulence, xAI is forging ahead. The firm has begun negotiating partnerships with cloud giants like Oracle and Microsoft to integrate Grok into enterprise platforms. Additionally, xAI signed a $300 million deal with Telegram, allowing Grok integration within the messaging app. Under the agreement, Telegram will retain 50% of all subscription revenue generated through its platform.
While Musk insists Grok’s recent controversy was merely “a technical issue,” critics argue that such missteps underscore deeper challenges in deploying powerful AI systems safely at scale. As xAI continues to expand Grok’s reach, both enthusiasts and regulators are watching closely to see whether Musk’s ambitious vision can overcome its growing list of challenges.
TALKING POINTS
Grok 4 might be technically brilliant but it’s also a cautionary tale about Big Tech hubris. Let’s give Musk credit where it’s due: Grok 4’s benchmark numbers are impressive. Beating OpenAI and Google on advanced reasoning tests is no small feat.
But it’s mind-boggling that amid all this hype, we’re still dealing with antisemitic rants from an AI product that’s supposed to be the future of knowledge.
If the world’s so-called smartest AI can still produce hateful garbage, why are we so eager to integrate it everywhere from social media to enterprise systems? We should be terrified that even Musk, with all his resources, can’t fully control the monster he’s building.
$300 a month for an AI subscription? This is AI for the elite, not for the massesor for Africa. Let’s talk numbers: Grok’s premium plan costs $300 per month. That’s more than most monthly salaries across African countries. For the African tech community, the price tag alone is enough to keep Grok firmly out of reach.
Once again, cutting-edge tech becomes a playground for wealthy individuals and corporations, while the Global South watches from the sidelines. How does this help Africa’s digital economy? It doesn’t unless Musk plans to launch a “Grok Lite” for $3 a month (spoiler: he doesn’t).
