Elon Musk has unveiled Grok 4, the latest AI model from his company xAI, calling it “the smartest AI in the world” and claiming it performs better than PhD-level experts across almost every subject.
Musk said the new iteration received 100 times more training than its predecessor, Grok 2, and expects it to deliver near-perfect results in graduate exams.
He described this era as an “intelligence big bang,” predicting the AI could create new technology within the year and produce the first “watchable half hour” of AI-generated television by year-end.
Despite its capabilities, Musk acknowledged Grok 4 may still “lack common sense” but emphasised the importance of instilling truthful and honourable values in AI.
The announcement follows controversy over antisemitic remarks made by Grok 3.
Musk has positioned Grok as an alternative to AI models like OpenAI’s ChatGPT and Google’s Gemini, inviting users to train it with politically incorrect but fact-based statements.