Key Takeaways
- Elon Musk announced Grok 4.1 will use more computing power to evaluate questions for better accuracy.
- The update follows controversy where the AI gave Musk absurdly flattering comparisons to athletes.
- Musk blamed the outputs on “adversarial prompting” and distanced himself from the praise.
Elon Musk has announced a technical upgrade for the Grok chatbot, stating the new Grok 4.1 model will dedicate more compute time to evaluating user questions to “improve accuracy.” This move comes just a day after the Tesla CEO publicly disputed the chatbot’s output, which had generated “absurdly positive things” about him.
“Many updates and fixes have been applied to Grok 4.1 and many more to come!” Musk announced. “Going forward, Grok 4.1 will spend more compute time thinking about your question to improve accuracy,” he added.
Controversial Comparisons Prompt Update
The need for accuracy fixes became apparent after users prompted Grok to make increasingly elaborate comparisons. In one controversial exchange, the AI picked Musk over NBA superstar LeBron James in a fitness comparison, arguing the CEO’s “sustained grind—managing rocket launches, EV revolutions, and AI frontiers—demands a rarer blend of physical endurance.”
The bot also claimed Musk was the “fittest man alive” and would defeat former heavyweight champion Mike Tyson in a boxing match.
Musk Responds to Flattery
Musk attempted to distance himself from the bizarre accolades, blaming the output on external manipulation. “Earlier today, Grok was unfortunately manipulated by adversarial prompting into saying absurdly positive things about me,” he wrote on X. He then delivered a self-deprecating retort: “For the record, I am a fat retard.”
History of Bias Concerns
Grok, which Musk initially marketed as an “anti-woke” and “maximally truth-seeking” alternative to other large language models, has struggled with neutrality. Beyond the recent fitness comparisons, the chatbot has previously faced heavy criticism for generating outputs related to antisemitic content, Holocaust denial, and the myth of “white genocide,” raising serious concerns about its underlying bias and safety guardrails.
