format(webp))
Zuck shrugs off DeepSeek, vows to spend hundreds of billions on AI
Meta remains committed to massive AI investments despite concerns over DeepSeek’s impact on GPU demand, which caused Nvidia’s stock to drop. CEO Mark Zuckerberg reaffirmed plans to spend heavily on AI infrastructure, seeing it as a long-term strategic advantage. Meta aims to make Llama 4 the leading AI model, surpassing both open-source and closed alternatives with advanced multimodal and agentic capabilities.
U.S. markets saw a sharp reaction on Monday amid speculation that DeepSeek’s AI models could reduce demand for GPUs, causing Nvidia’s stock to plunge nearly 20%.
Despite this, Meta remains committed to AI investment. During the company’s first-quarter earnings call on Wednesday, CEO Mark Zuckerberg reaffirmed plans to spend aggressively on AI, potentially reaching “hundreds of billions of dollars” over time. Meta has already allocated over $60 billion for capital expenditures in 2025, primarily for data centers.
Addressing concerns about DeepSeek’s impact on AI spending, Zuckerberg emphasized that investing in AI infrastructure remains a key strategic advantage. While acknowledging DeepSeek as a competitor, he noted it is still too early to determine if GPU demand will decline, as chips remain essential for inference tasks given Meta’s massive user base.
Zuckerberg stressed that building AI infrastructure at scale would continue to enhance both service quality and reach. He also outlined Meta’s ambition for its next AI model, Llama 4, aiming to surpass both open-source and closed models like ChatGPT. Unlike Llama 3, which focused on making open-source AI competitive, Llama 4 is designed to lead the field, incorporating advanced multimodal and agentic capabilities.
What's Your Reaction?
