Investing.com — Shares of AI infrastructure companies plummeted on Monday as investors responded to news that China’s DeepSeek has developed a highly efficient open-source language model, potentially impacting future capital expenditures in the AI industry.
The innovative model, which rivals the performance of GPT-4 with significantly less computing power, has introduced uncertainty regarding the growth of demand for computing power.
The architecture of DeepSeek’s model, which combines Mixture of Experts (MoE) and Meta-Learning Algorithms (MLA) with high-quality parameter processing, has set a new precedent for model efficiency. This development could prompt the AI sector to prioritize return on investment (ROI) more heavily, leading to a reassessment of necessary computing power.
As a result, projections for AI-related capital expenditures in 2026 are now being questioned, with potential declines or stagnation on the horizon.
GE Vernova, Constellation Energy (NASDAQ:) Corp., Vistra Corp., and Ciena (NYSE:) Corp. all experiencing significant drops in their stock prices, with Vertiv shares down as much as 14%.
Jefferies analyst Edison Lee commented on the situation, stating, “The market naturally will worry about demand growth in computing power.”
This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.