The tech world, including Wall Street, took notice last month as DeepSeek burst onto the scene with its high-performing AI model, R1. Experts believe this is just the beginning of the Chinese tech startup’s potential impact on the artificial intelligence field.
DeepSeek grabbed headlines in late January with its R1 AI model, which some experts say nearly matches the performance of OpenAI’s comparable model but at a significantly lower cost. The news caused tech stocks to tumble briefly as DeepSeek even briefly unseated ChatGPT to become the top app in Apple’s App Store. This achievement has prompted tech giants in the US to question America’s position in the global AI race and the significant financial investment that competition requires.
While Vice President JD Vance didn’t mention DeepSeek or China directly in his remarks at the Artificial Intelligence Action Summit in Paris, he underscored the importance of the United States leading the sector. “The United States of America is the leader in AI, and our administration plans to keep it that way,” he said, while also noting that “America wants to partner” with other countries.
What makes DeepSeek so impactful extends beyond its efficiency and power. Experts point to DeepSeek R1’s ability to reason and “think” through answers, providing quality results, alongside the company’s decision to make parts of its technology publicly accessible. This openness will push the field forward.
Artificial intelligence has become a flashpoint in the last two years. The recent rise of ChatGPT and other generative AI services has reshaped the way people work, communicate, and find information. This has created Wall Street darlings and upended the trajectory of some of Silicon Valley’s top companies. So any development that allows for more efficient and capable models is something to be watched.
“This is definitely not hype,” said Oren Etzioni, former CEO of the Allen Institute for Artificial Intelligence. “But also, this is a very fast-moving world.”
AI’s TikTok Moment
Tech leaders have been quick to respond to DeepSeek’s success.
Google DeepMind CEO Demis Hassabis called any hype around DeepSeek “exaggerated,” but also said its model as “probably the best work I’ve seen come out of China,” according to CNBC. Microsoft CEO Satya Nadella mentioned during the company’s quarterly earnings call in January that DeepSeek has some “real innovations.” Apple CEO Tim Cook said that “innovation that drives efficiency is a good thing,” during the iPhone maker’s earnings call.
However, not all reactions have been positive. Semiconductor researcher SemiAnalysis questioned DeepSeek’s claim that training the model cost only $5.6 million. OpenAI told The Financial Times that it found evidence that DeepSeek used the US company’s models to train its own. “We are aware of and reviewing indications that DeepSeek may have inappropriately distilled our models, and will share information as we know more,” an OpenAI spokesperson said in a comment to CNN. DeepSeek has not yet responded to requests for comment.
Additionally, a pair of US lawmakers has already called for the technology to be banned from government devices. This came after security researchers pointed out its potential links to the Chinese government, as the Associated Press and ABC News reported. Similar concerns have been raised about the popular social media app TikTok, which must be sold to an American owner or face a ban in the US.
“DeepSeek is the TikTok of (large language models),” Etzioni said.
DeepSeek’s Deep Impression
Tech giants are already considering how DeepSeek’s technology could influence their products and services. “What DeepSeek gave us was essentially the recipe in the form of a tech report, but they didn’t give us the extra missing parts,” said Lewis Tunstall, a senior research scientist at Hugging Face, an AI platform that offers tools for developers. Tunstall is leading an effort at Hugging Face to fully open source DeepSeek’s R1 model. While DeepSeek released a research paper and the model’s parameters, it didn’t reveal the code or training data.
Nadella stated on Microsoft’s earnings call that Windows Copilot+ PCs, or PCs built to certain specs to support AI models, should be able to run AI models derived from DeepSeek R1 locally.
Mobile chipmaker Qualcomm announced that models derived from DeepSeek R1 could run on smartphones and PCs powered by its chips within a week.
AI researchers, academics, and developers are still exploring what DeepSeek means for the advancement of AI. DeepSeek’s model isn’t the first to have open-source elements. Neither is it the first to be able to reason over answers before responding; OpenAI’s o1 model from last year could do that, too. The significance of DeepSeek comes from its reasoning abilities and its capacity to learn from other models, as well as the public nature of its inner workings. Those who use DeepSeek’s R1 model can also see its “thought” process as it generates answers to questions.
“You can see the wheels turning inside the machine,” Durga Malladi, senior vice president and general manager for technology planning and edge solutions at Qualcomm, said to CNN.
Tunstall suggests that a wave of new, similarly reasoning models may be available in the not-too-distant future. That could be critical as tech giants race to develop AI agents, which many in Silicon Valley see as the next evolution of chatbots. While that shift hasn’t fully happened yet, Grok 3, the next iteration of the chatbot on the social media platform X, will have “very powerful reasoning capabilities,” according to its owner, Elon Musk, during a video appearance at the World Governments Summit.
For now, those in the AI industry will continue to experiment with what DeepSeek has to offer, until the next major breakthrough emerges.
“I certainly predict that in the next 12 months, it’ll be supplanted by something else,” said Etzioni. “But it’s a very real advance.”
