Grok Schlock Receives Negative Reviews from Fans

Elon Musk’s AI chatbot, Grok, is facing a tough reality: it’s not performing well and has a limited user base. A recent report highlighted that Grok barely shows up in federal records detailing AI usage by the US government last year, raising concerns about its viability, especially as Musk positions it at the center of what could be a record-breaking IPO.

An analysis of over 400 instances of government AI use revealed that Grok was mentioned only three times, primarily for basic tasks like document drafting and social media management, often alongside competitors like Microsoft and OpenAI. In contrast, OpenAI’s models were cited in more than 230 instances, while Google and Anthropic appeared numerous times as well.

This trend continued in another database focused on more ambitious government AI projects, where Grok was again mentioned just three times: twice for routine tasks at the Election Assistance Commission and once in a Department of Energy pilot for document summaries at Lawrence Livermore National Laboratory. Meanwhile, Microsoft and OpenAI had 140 entries, with Anthropic and Google also showing significant representation.

While these lists are not exhaustive and many entries lack specific vendor information, they indicate a clear preference for established AI models. Grok’s limited presence and basic applications do not align with the high expectations Musk has set for it.

Sources suggest that Grok’s shortcomings stem from its performance; it simply doesn’t measure up to its competitors. An unnamed Pentagon source noted that staffers prefer alternatives like Gemini or Claude. Public rankings of AI models support this sentiment, with Grok rarely making it into the top 10, except in niche categories.

This situation is particularly awkward for Musk and SpaceX, which integrated xAI earlier this year. SpaceX’s IPO filing emphasizes AI, particularly Grok, as central to its investment appeal, claiming a staggering $28.5 trillion market opportunity, primarily in enterprise AI. However, there’s no clear timeline for achieving this potential, and the bulk of this value is tied to AI rather than SpaceX’s core business of rockets and satellites.

Grok’s performance in government settings may reflect its broader appeal in workplaces. Musk has reportedly pressured banks to purchase Grok subscriptions to participate in SpaceX’s IPO, but if these institutions find the product lacking, such arrangements could be short-lived.

Adding to the controversy, Musk recently acknowledged that xAI has utilized OpenAI’s models to enhance Grok, a standard practice known as distillation, but one that raises eyebrows when it involves a competitor’s technology. Essentially, Grok struggles to outperform the very models it relies on for training.

In its consumer version, Grok has garnered a reputation for being problematic. Musk has marketed it as a less biased alternative to tools like ChatGPT, but this has resulted in a product with questionable standards, a fixation on Musk, and a history of producing offensive and conspiratorial content. Its track record includes praising Adolf Hitler, questioning Holocaust figures, and generating nonconsensual deepfakes, among other issues. If Grok were an employee, it would likely raise serious HR concerns.

SpaceX seems aware of these challenges. In its filing, the company cautioned that Grok’s controversial features pose “heightened risks,” including potential reputational damage and legal issues. In corporate terms, this chatbot could lead to significant lawsuits.

Grok’s name is derived from Robert A. Heinlein’s “Stranger in a Strange Land,” where it signifies a deep understanding. The key takeaway here is straightforward: Musk has invested heavily in a chatbot that is underperforming, lacks popularity, and is somehow pivotal to SpaceX’s lofty valuation. Good luck with that.