Key Takeaways
- Elon Musk declared that xAI’s upcoming Grok 5 model represents his target for achieving artificial general intelligence
- The company’s Grok 4.8 iteration, featuring 2.5 trillion parameters built on a novel C++ architecture, will complete its training cycle this week before transitioning to reinforcement learning
- In performance comparisons, Musk positioned Grok 4.7 alongside Anthropic’s Opus 5.0, describing the two as comparable in aggregate capabilities
- Despite his AGI ambitions, Musk endorsed statements from OpenAI and Anthropic leadership advocating for measured AI advancement on the same date
- Semiconductor manufacturers Nvidia and AMD experienced significant stock declines following industry warnings about AI development pace
On September 14, Elon Musk declared that xAI’s forthcoming Grok 5 represents the iteration he anticipates will cross the threshold into artificial general intelligence.
The announcement arrived via two separate messages posted to X, immediately sparking conversations across research communities and financial markets.
The term AGI describes artificial intelligence systems capable of performing at or above human levels across diverse cognitive tasks, as opposed to specialized applications designed for narrow functions.
xAI’s Development Timeline Revealed
In earlier communications on the same day, Musk indicated that Grok 4.8 would conclude its primary training phase before the week’s end. This iteration incorporates 2.5 trillion parameters and operates on a proprietary C++ infrastructure created internally by xAI.
Following the completion of initial training, Grok 4.8 will transition into reinforcement learning phases, where the system refines its capabilities through structured feedback mechanisms.
When questioned about Grok 4.8’s proximity to AGI capabilities, Musk responded directly: “That will be Grok 5.”
He subsequently provided additional context regarding xAI’s progression strategy. According to Musk, Grok 4.7 matches Anthropic’s Opus 5 in general capability, with each system demonstrating advantages in different domains. Grok 4.9 would target performance levels comparable to Astra or Fable. Regarding Grok 5, Musk stated: “Maybe better than anything. We shall see.”
The company’s rapid advancement stems from integrating massive computational infrastructure with live data streams from X and technical expertise transferred from SpaceX operations.
Safety Discourse and Financial Impact
Musk’s AGI declarations coincided with his public endorsement of statements from Dario Amodei of Anthropic and Sam Altman of OpenAI urging restraint in accelerating AI capabilities.
This dual positioning attracted scrutiny. While pursuing AGI development aggressively, Musk simultaneously validated concerns about the technology’s potential risks.
Market responses to safety discussions among AI executives impacted semiconductor equities negatively. Nvidia shares declined 3.46% while AMD experienced a 6.24% drop as investors weighed implications for AI chip demand.
Industry observers highlight several risk categories: AI systems potentially optimizing for objectives misaligned with human welfare, consolidation of technological power among limited corporate entities, and unpredictable advances in system capabilities.
Musk has maintained that advanced AI systems could dramatically increase worldwide economic productivity in coming years, accelerating progress in software development, robotics applications, and scientific investigation.
Whether Grok 5 achieves the AGI milestone Musk envisions remains speculative. While xAI has rapidly narrowed performance differences with more established laboratories, creating a system that matches human-level competence across all cognitive domains would represent an unprecedented achievement.
Grok 4.8 is projected to finish training and begin reinforcement learning operations before the current week concludes.





