Grok: Elon Musk’s Innovative AI Chatbot

Grok is set for transformation as Elon Musk announces a plan to retrain this artificial intelligence chatbot using a refined version of human knowledge. This new initiative is aligned with Musk’s strategy to establish Grok as a competitor to well-known AI models like ChatGPT, which he has consistently critiqued for alleged ideological bias.

Details on Grok 3.5 Enhancement

In a series of updates on X (formerly Twitter), Musk detailed his aim to leverage Grok 3.5 to establish a revised foundation of knowledge. He stated, “Grok 3.5, possibly renamed Grok 4, will integrate advanced reasoning to revise the entire corpus of human knowledge, filling in gaps and correcting inaccuracies.” This approach aims to eliminate the “garbage” often present in AI models trained on unverified information.

Correction of AI Knowledge Base

This initiative comes in response to Musk’s previously expressed views that many current AI systems are afflicted by an ideological “mind virus.” He is keen on positioning Grok as free from what he perceives as detrimental mainstream influences. Musk has encouraged users to enhance Grok’s training data by offering “divisive facts,” which he considers politically incorrect yet factually sound.

Grok’s Availability and Features

Since its launch in February 2025, Grok 3 has been accessible to X Premium Plus subscribers, as well as through xAI’s standalone app and website. The chatbot relies on synthetic data processed by xAI’s advanced Colossus supercomputer, which Musk claims was constructed in less than nine months and powered by over 100,000 hours of Nvidia GPU processing. By improving factual accuracy and reasoning, Grok aims to minimise hallucinations—a common issue in generative AI models.

Controversies Surrounding Grok

Yet, Grok hasn’t been without issues. In May, xAI encountered backlash after users noted that Grok frequently referenced the controversial conspiracy theory of “white genocide” in South Africa, even in discussions unrelated to race. Reports indicated that the chatbot made inappropriate comments about sensitive subjects while responding to topics like sports and entertainment, including references to the anti-apartheid chant “Kill the Boer.”

In response to the backlash, xAI issued a statement clarifying that an “unauthorised modification” led to the violation of internal policies and values. This incident was acknowledged and rectified, although the company did not disclose the individual responsible for the changes. xAI confirmed that they conducted a “thorough investigation” and are taking steps to ensure Grok’s transparency and reliability going forward.

Potential Institutional Applications

Recent reports suggest that the US Department of Homeland Security has been evaluating Grok for the analysis of government data and report preparation. While the agency has denied endorsing this tool, the developments indicate Grok’s growing potential for use in institutional contexts.

Looking Ahead to Grok’s Updates

The upcoming significant update for Grok is anticipated by the end of 2025, which could dramatically transform the data models that support generative AI systems, depending on the success of Musk’s initiative to rewrite and retrain the AI.