Microsoft CEO Satya Nadella praised its “real innovations.” OpenAI’s Sam Altman lauded it as “clearly a great model.” Apple CEO Tim Cook stated that “innovation that drives efficiency is a good thing.” Palantir’s Alex Karp emphasized the importance of “an all-country effort.”
The tech CEOs were discussing China’s DeepSeek, which emerged from obscurity to the forefront of the tech industry this week. Over the past few days, these executives and many others have been addressing inquiries about the startup lab’s groundbreaking artificial intelligence model, which has impressed experts and was reportedly more cost-effective to develop compared to similar models in the U.S.
DeepSeek’s mobile app quickly rose to the top of Apple’s App Store rankings early in the week and maintained its leading position through Friday, surpassing OpenAI’s ChatGPT. Reports that its new R1 model, competing with OpenAI’s o1, was created for just $6 million caused shares of chipmakers Nvidia and Broadcom to plummet by 17% on Monday, resulting in a combined $800 billion loss in market capitalization.
The timing was significant as DeepSeek’s launch coincided with the beginning of tech earnings season, with major companies like Meta, Microsoft, Tesla, and Apple reporting their earnings between Wednesday and Thursday. This occurred during the first week of President Donald Trump’s second term, who has stressed the importance of the U.S. leading in AI, especially against China. Trump announced a project named Stargate in his first week back in office, urging OpenAI, Oracle, and SoftBank to invest billions to enhance domestic AI infrastructure.
There has been extensive online discussion regarding the impact of DeepSeek’s launch and the validity of its financial success. A report from research firm SemiAnalysis estimated DeepSeek’s costs to be “well higher than $500M over the company history.”
AI is currently a focal point for every company, particularly in the technology sector, where industry leaders are investing billions in constructing data centers and acquiring advanced chips to develop more potent models. For U.S. tech leaders, the common goal is to outperform their primary competitor, China, in the future-defining technology.
Karp, Palantir’s CEO, cautioned in an interview with CNBC’s Sara Eisen that while technology isn’t inherently good and could pose risks, it necessitates a collective effort to stay ahead. Cook, during Apple’s earnings call, was asked about the impact of DeepSeek on the company’s margins and computing costs. Cook highlighted the benefits of innovation driving efficiency and expressed confidence in their integrated approach to silicon and software.
Zuckerberg, Meta’s CEO, suggested that the situation is complex and that early reports on a single model don’t fundamentally alter the landscape. Meta announced plans to invest between $60 billion and $65 billion by 2025 to expand its AI-related computing infrastructure.
Nadella mentioned during Microsoft’s call that DeepSeek R1 will soon be compatible with their Copilot+ PCs and GPUs on Windows. He noted that AI advancements are following a similar trajectory as regular computing cycles, emphasizing the importance of continuous improvement and customer benefits.
Hoffman, a LinkedIn co-founder and early investor in OpenAI, reassured that the competition is intensifying but doesn’t signify a loss for American technology. Altman, OpenAI’s CEO, praised DeepSeek’s R1 model and stressed the need for democratic AI to succeed, echoing sentiments shared by President Trump.
Altman affirmed OpenAI’s commitment to advancing their models, expressing enthusiasm for healthy competition. The report was contributed to by CNBC’s Jordan Novet, Hayden Field, Kif Leswing, Ashley Capoot, and Jonathan Vanian.
