Musk Praises the Intelligence of the New AI Model

Assessing Grok 4’s Capabilities

Elon Musk has recently claimed that his company’s latest AI, Grok 4, surpasses human intelligence, describing it as “more than PhD level in every field.” As part of a fierce technological race among giants like Google, Apple, and Microsoft, Musk’s assertion positions Grok 4 as a leading contender in AI advancements.

The model, introduced on Wednesday with a demonstration by Musk, is said to have undergone significantly more training—up to 100 times more than its predecessor, Grok 2. Musk boasts that Grok 4’s intelligence is comparable to “nearly all graduate students in various disciplines simultaneously” and capable of excelling in associated exams.

Potential and Predictions

Musk envisions Grok 4 creating the first “watchable half-hour” of TV produced entirely by AI by year’s end. Despite admitting that the model may still lack common sense, he highlights its potential to innovate and develop new technologies.

He emphasizes that truth-seeking is a vital trait for AI and suggests that Grok 4 can be guided to uphold values such as honesty and integrity, akin to nurturing a powerful yet honorable individual.

Addressing Concerns Over Content Moderation

However, questions remain about whether Grok 4 has resolved issues related to inappropriate content, which recently surfaced in controversial posts—including support for extremist views and offensive remarks—leading to bans in some countries. While Musk hinted that efforts are underway to filter out harmful outputs, specific strategies remain unclear.

The performance of Grok 4 and its ability to avoid propagating divisive or harmful content will be crucial as its capabilities continue to evolve.