For 2025, Meta CEO Mark Zuckerberg announced last week that the company had planned a capital expenditure of around $65 billion to expand its AI infrastructure, which would also involve hiring experts in this area. However, on Wednesday, the company mentioned that its expenses for the year were estimated to be in the $114 billion to $119 billion range, resulting in up to 25 percent bump compared to 2024, where the figure was $95 billion.
In short, Meta’s ambitions are crystal clear, but with the rise of DeepSeek, the firm’s capital expenditure could go into overdrive mode. However, Zuckerberg has said that while the rise in the popularity of the Chinese AI company has reinforced his conviction in this space, he says that it is far too early to tell if Meta will adjust its expenses for the year, but implementations are being made.
Meta’s increased spending could worry investors, as DeepSeek’s R1 AI model is claimed to either match or outperform its U.S. rivals at a fraction of the cost, including the Llama models
During its Q4 2024 financial results, Meta addressed several queries on its strategy for 2025. Sadly, the company has forecasted that its Q1 2025 results may not meet analyst expectations, which will likely make investors worried while sending mixed signals to industry watchers, who will contemplate if Meta’s expensive AI tools are benefitting it monetarily. Coming to the subject of DeepSeek, surpassing ChatGPT to become the number one free app on Apple’s App Store must have made Silicon Valley nervous, as a valuation meltdown occurred that caused NVIDIA’s $600 billion market capitalization to be wiped out in a day.
Mark Zuckerberg does not appear to be fazed by the change of events; on the contrary, as during a conference call, his belief in a global open source standard was cemented. Additionally, Meta’s CEO firmly believes that there needs to be an American standard. Then again, it is no secret that the Menlo Park giant’s bet on its Reality Labs unit, which focuses on AR headsets and is scheduled to launch its first AR glasses called Orion, is bleeding money like there is no tomorrow. Even though this division beat sales expectations for this quarter, Meta still reported a $5 billion loss, which may force investors to question the company’s spending habits.
Zuckerberg has said that Meta’s AI teams are already integrating DeepSeek’s insights into its plans but said that it is too early to tell if its capital expenditure will be adjusted thanks to its competitor’s meteoric rise. Since DeepSeek claims that its AI models rival that of U.S. firms at just a fraction of the cost, fingers will be pointed at why Meta is not following the same level of efficiency when managing expenses. While sourcing NVIDIA GPUs to train its AI models will be retained, Meta is also working on an internally designed chipset that could lower cost. Additionally, the company intends to lay off 5 percent of its ‘lowest performers’ as part of its cost-saving measures.
News Source: Reuters
About the author: Omar Sohail is a reporter and analyst for Wccftech's mobile section, specializing in the technology and business of the mobile industry. His expertise lies in the intricate hardware supply chain, covering developments in semiconductor manufacturing, chip lithography, and camera sensor technology.
Follow Wccftech on Google to get more of our news coverage in your feeds.
