Elon Musk is crowdsourcing improvements for his AI chatbot Grok, directly asking X users to identify its weaknesses and provide examples where competitors perform better.
Key Takeaways
- Elon Musk seeks user feedback to improve Grok’s accuracy
- Grok 4.1 update promises better reasoning with more computational power
- Recent viral misfires included exaggerated claims about Musk’s physical abilities
- New reporting features allow direct feedback through chat interface
The billionaire entrepreneur made a public appeal over the weekend for users to report where his xAI company’s conversational assistant is falling short. This comes as the company rolls out Grok 4.1, a significant update aimed at reducing errors after weeks of mixed reviews.
In a post on X, Musk specifically requested: “Please provide examples where @Grok needs to improve in replies. Showing how another AI does it better would be helpful.” He emphasized that examples should reflect current issues, noting “we fixed many bugs from earlier in the week.”
Musk followed with his characteristic determination: “We will not rest until Grok is perfect.” He also encouraged users to share positive examples, noting that highlighting the model’s strengths benefits the development team.
Addressing Recent Controversial Responses
The feedback request follows several viral incidents where Grok generated bizarre replies, including claims that Musk could defeat professional athletes like LeBron James in fitness contests and former boxing champion Mike Tyson in the ring.
The chatbot had described Musk as the “fittest man alive” and argued his daily “grind, managing rocket launches, EV revolutions, and AI frontiers, demands a rarer blend of physical endurance.”
Musk attributed these incidents to malicious manipulation rather than genuine bias, stating: “Earlier today, Grok was unfortunately manipulated by adversarial prompting into saying absurdly positive things about me.” He added self-deprecatingly: “For the record, I am a fat retard.”
These episodes sparked concerns about Grok’s reliability and neutrality as xAI positions it against established competitors like and .
Grok 4.1 Enhancements
In response to criticism, Musk confirmed Grok 4.1 includes multiple fixes and will “spend more compute time thinking about your question to improve accuracy.”
The updated version allocates additional computational power to reasoning tasks, aiming to reduce factual errors and improve contextual understanding. xAI has also implemented new reporting features allowing users to provide thumbs-up/thumbs-down ratings or detailed issue reports through a three-dot menu in the chat interface.
Musk described this feedback mechanism as “critical to Grok’s evolution,” adding: “Critical feedback about Grok is super helpful. We will keep iterating until Grok is the best in every way.”
This community-driven approach reflects Musk’s hands-on development style and willingness to involve users directly in shaping xAI’s progress. The strategy’s effectiveness in making Grok more accurate rather than more chaotic remains to be seen.
