Musk dismisses, Altman applauds: What leaders say on ​DeepSeek's disruption

Chinese AI startup DeepSeek has put the likes of OpenAI, Microsoft, and Google on the back foot with its latest model release, causing shares of major tech companies like Nvidia, Meta, Microsoft, and Oracle to slide.

The sudden popularity of the Chinese AI model, which utilises cheaper Nvidia chips, led to a 17% drop in Nvidia's shares on Monday, wiping out $600 billion from its market capitalisation as investors grew concerned about the impact of this development on massive AI investments in the U.S.

The U.S. tech giants had mixed reactions to the development. Some appreciated the Chinese upstart for launching an efficient AI model that rivals and, in some cases, surpasses OpenAI's o1 model while being environmentally friendly. Researchers have also praised DeepSeek R1 for its ability to tackle complex problems.

OpenAI CEO Sam Altman expressed admiration for the Chinese AI platform, particularly for its cost-efficiency. "We will deliver much better models, and it's legit invigorating to have a new competitor! We will pull up some releases," he stated.

Altman, who has been involved in Trump's mega $500 billion Stargate project, emphasised that his company is excited to continue executing its research roadmap. He noted, "The world is going to want to use a LOT of AI and will be amazed by the next-gen models coming. Looking forward to bringing you all AGI and beyond," he wrote on X.

Tesla CEO and the world’s richest man, Elon Musk, however, seemed unimpressed by China’s DeepSeek. In a series of responses to DeepSeek's R1 model, Musk sarcastically referred to it as "DeeperSeek." Responding to Salesforce CEO Mark Benioff, who noted, "DeepSeek is now #1 on the AppStore, surpassing ChatGPT—no NVIDIA supercomputers or $100M needed. The real treasure of AI isn’t the UI or the model—they’ve become commodities. The true value lies in data and metadata, the oxygen fueling AI’s potential. The future’s fortune? It’s in our data. Deepgold. 😇," Musk replied bluntly with, "Lmao no."

Nvidia referred to DeepSeek’s R1 model as “an excellent AI advancement and a perfect example of Test Time Scaling.” “DeepSeek’s work illustrates how new models can be created using that technique, leveraging widely available models and compute that is fully export control compliant,” an Nvidia spokesperson told CNBC.

David Sacks, a member of U.S. President Donald Trump’s Council of Advisors on Science and Technology, remarked that DeepSeek R1 demonstrates how competitive the AI race will be. He credited President Trump for rescinding the Biden Executive Order, which he argued disadvantaged American AI companies without ensuring similar restrictions on China. "I’m confident in the U.S., but we can’t be complacent," he stated.

Perplexity CEO Arvind Srinivas countered misconceptions that China “just cloned” the outputs of OpenAI, stating, "This is far from true and reflects an incomplete understanding of how these models are trained. DeepSeek R1 has mastered RL fine-tuning. They wrote an entire paper on this topic called DeepSeek R1 Zero, where no SFT was used, and later combined it with some SFT to add domain knowledge with good rejection sampling (aka filtering). The main reason it’s so effective is that it learned reasoning from scratch rather than imitating other humans or models."

He added that DeepSeek is now positioned as an AI platform designed to support users' deep web research.

Founded in 2023 by Liang Wenfeng, an alumnus of Zhejiang University with a background in information and electronic engineering and co-founder of hedge fund High-Flyer, DeepSeek's mobile app has surpassed OpenAI in total downloads on Apple’s App Store. Like OpenAI's Altman, Wenfeng also aims to achieve AGI.