In business, profit is the ultimate goal—but the methods of achieving it vary from one company to another.
Since the launch of OpenAI’s ChatGPT, Large Language Models (LLMs) have become the center of the tech world. Companies, both large and small, have rushed to either compete or integrate AI into their platforms, fueling a race for dominance in the West.
But China was watching—and it wasn’t long before it joined the race.
It's only then, that DeepSeek joined the saturated crowd.
Its emergence has sent shockwaves through Silicon Valley, stirring unease among tech giants. Even the U.S. government has taken notice, raising concerns over China's sudden rise in the AI arena.

DeepSeek, a Chinese AI startup just over a year old, has disrupted the tech world, drawing both praise and apprehension.
While China has previously developed LLMs capable of competing with Western counterparts, DeepSeek has achieved similar success—but cheaper to begin with.
In particular, the product called the DeepSeek-R1, has not only matched but, in also outperformed the top-of-the-line tools from OpenAI, Meta and others on various third-party benchmarks designed to test AI performance across different subjects.
With its ability to combine real-time data with reasoning, it's able to beat OpenAI o1 in some benchmarks
Besides being more or less on par with advanced models from American tech giants — and it was also built using way less resources, trained using less advanced and number of GPU chips (due to strict U.S. export restrictions on advanced chips to China), and demands far less data center power to run.
Whereas world’s leading chatbots that the notion that cutting-edge AI requires massive resources, DeepSeek managed to compete on the same playing field with those from the West, but without too much spending.
DeepSeek was founded with a capital of 10 million yuan ($1.4 million), and that reportedly trained R1 at a fraction of the cost, reportedly around $5 million.
In comparison, Meta and Microsoft have committed to capex of $65 billion or more this year, largely on AI infrastructure.

Unlike many Western-made LLMs, which are proprietary and have their full capabilities locked behind paywalls, DeepSeek-R1 was released as a fully open-source model.
This openness has allowed developers and researchers to fine-tune DeepSeek-R1 for a wide range of specialized tasks—whether optimizing it for mobile devices or integrating it with other AI models.
Even for developers, DeepSeek’s API pricing is over 90% cheaper than OpenAI’s equivalent o1 model, making it an attractive alternative.
As a result, DeepSeek-R1 has rapidly gained traction on platforms like Hugging Face, where it has become one of the most downloaded and actively modified AI models.
In many ways, DeepSeek’s emergence challenges the belief that AI progress demands ever-increasing computational power and energy—proving that high performance can be achieved with fewer resources.
But DeepSeek isn’t just for developers and AI researchers.
The company offers a free website and mobile app, available even in the U.S., featuring an R1-powered chatbot with an interface nearly identical to OpenAI’s ChatGPT.

DeepSeek’s impact on the global tech industry and supply chain is profound.
Global technology stocks tumbled on January 27 as hype around DeepSeek’s innovation have made investors to worry about their investments in AI hardware suppliers such as Nvidia Corporation.
The greater efficiency of the model puts into question the need for vast expenditures of capital to acquire the latest and most powerful AI accelerators, a market that has long been dominated by Nvidia.
As investors offloaded their Nvidia stock in response, the company's shares went down 17% that day, instantly eliminating $589 billion of value, a moment considered a stock market record.
Not just Nvidia, because some other semiconductor machine maker ASML Holding NV and other companies that also benefited from booming demand for cutting-edge AI hardware are also tumbling.
DeepSeek's emergence also put extra attention on U.S. export curbs of such advanced semiconductors to China, with people asking whether it's even necessary.
The curb, part of which is intended to prevent a breakthrough of the sort that DeepSeek appears to represent.

Meta that heard about this R1's unique achievements, even assembled four war rooms of engineers to determine how this Chinese hedge fund managed to release an AI game-changer that may already rival its own technology,
With the company's team on high alert, Meta AI infrastructure director Mathew Oldham has reportedly told colleagues that DeepSeek’s newest model could outperform even the next version of Meta’s LLama AI, which Zuckerberg said could be released in “early 2025,”
Of the four war rooms Meta has created to respond to DeepSeek’s potential breakthrough.
The first two teams shall try to decipher how High-Flyer lowered the cost of training and running DeepSeek with the goal of using those tactics for LLaMA, reports said, citing one anonymous Meta employee.
As for the remaining two teams, one will try to find out which data DeepSeek used to train its R1 model, and the other will consider how the company can restructure LLaMA models based on attributes of the DeepSeek models.
Founded in 2023 by Liang Wenfeng, the head of AI-driven quant hedge fund High-Flyer, DeepSeek is based in Hangzhou and specializes in developing open-source AI models.
Wenfeng, born in 1985, is an engineering graduate Liang, but has never studied or worked outside of mainland China.
He received bachelor’s and masters’ degrees in electronic and information engineering from Zhejiang University.
With just $1.4 million in registered capital, according to company database Tianyancha, Wenfeng bottleneck isn't about fundraising.
DeepSeek initial struggle to advance was due to the U.S restriction, which prevents companies in China from buying America's best computer chips. What's more, Wenfeng also said that most of his top researchers were fresh graduates, despite graduating from top Chinese universities.
"More investment does not necessarily lead to more innovation. Otherwise, large companies would take over all innovation," he said.
Since the rise of DeepSeek, Wenfeng has been compared to OpenAI founder Sam Altman. But unlike Altman, the Chinese citizen keeps his profile low, and seldom speaks publicly.
.

What's initially known is that, DeepSeek's R1 is like all other Chinese AI models: it adheres to China's censorship.
The R1 can self-censor on topics deemed sensitive in China, which includes things like queries about the 1989 Tiananmen Square protests or geopolitically fraught questions such as the possibility of China invading Taiwan.
R1 is capable of delivering detailed responses about political figures and such, but refrains from speaking much about Chinese President Xi Jinping.
But the R1 literally came out of nowhere, and disrupted the landscape dominated by Americans, the moment the R1 hits the market.
It sent shockwaves throughout Silicon Valley and far beyond.
And this, many people have become well aware that the U.S. is in an AI arms race with China.
This is the moment where the U.S. is questioning itself, thinking whether it's really a 'Sputnik Moment,' which refers to the Soviet Union’s launch of the first Earth-orbiting artificial satellite Sputnik 1 in 1957 which caught the US unprepared, or a 'Potemkin Moment.' which refers to a pretentiously showy façade intended to mask or divert attention from an embarrassing or shabby fact or condition.
Either way, DeepSeek is in the headlines of tech, and it's having its 'TikTok Moment.'