DeepSeek's Surge: China's AI Breakthrough Disrupts Global Tech and Challenges US Leadership in the AI Market

A 17-year-old boy name Liang Wenfeng, from the Guangdong Province of China in the late nineties pursued his studies while the market was focused on business and valued business over education. He completed his master's degree in Information and Communication Engineering from Zhejiang University in 2010.

Now reach back in January 20, 2025, DeepSeek, a newly launched AI application from China, has created significant turbulence in the global AI landscape. DeepSeek App  is making waves, going viral and securing the top spot as the #1 AI app in the app stores. Its rapid success is a clear sign of its impressive performance and the growing demand for cutting-edge AI technology. 

It triggered a disruption in the US market, wiping out USD 600 billion from major corporations in the stock market within just an hour. The challenge was taken on by a small team from a little-known company led by Liang Wenfeng, who has emerged as a key figure in this unprecedented event.

In a recent statement, US President Donald Trump commented on DeepSeek's breakthrough, describing it as a "positive thing" and a wake-up call to the US IT market, urging the US to accelerate its own advancements in AI technology.

Liang Wenfeng, now the 39 year old founder of DeepSeek, established the company in 2023. In just a matter of weeks, he has become a prominent figure in China’s tech sector, embodying the nation’s aspirations to navigate the increasing challenges posed by US export restrictions.

Despite his typically reserved nature and keeping an extremely low profile until the launch of DeepSeek in January 2025, Liang was invited last year to speak at an exclusive symposium hosted by China’s Premier Li Qiang. He was one of nine individuals asked to give a speech at the prestigious event.

He was also featured in a handful of interviews in recent years, where he expressed his desire to tackle global challenges with the best AI technology. “China’s AI can’t be in the position of following forever,” Liang said, addressing not just the time gap but the need for fundamental innovation. He pointed out that while China is often seen as just one or two years behind the United States, the real difference lies in the ability to create original solutions, rather than merely replicating what has already been done elsewhere.

What is DeepSeek?
DeepSeek is a newly launched AI app from a Chinese startup, designed to compete with MEGA7 like OpenAI, Meta or other in the line and is capable of performing at an equivalent or superior level to their technologies.

MEGA7 : Mega7 refers to the seven tech giants of the world: 1. Microsoft, 2. Apple, 3. Google (Alphabet), 4. Amazon, 5. Meta (Facebook), 6. Tesla, and 7. Nvidia.

Key Facts and Features of DeepSeek:

• Open Source: Like its competitors, DeepSeek is open-source, allowing anyone to access its code and deploy it on local computing resources. This flexibility makes it easy to run and customize for various applications without needing special infrastructure.
   
• Pocket Friendly : The cost of training the DeepSeek model is significantly lower than that of market giants, making it a more cost-effective solution for advanced AI tasks.
   
• Machine of Experts (MoE) Architecture: It uses different specialized models, called "experts," to handle different parts of a task. A system picks the best expert for each job.
   
• Selective Parameter Activation: Only about 37 billion out of 671 billion parameters are activated at a time, optimizing computational resources and reducing costs without sacrificing performance.
   
• Pure Reinforcement Learning: The platform employs Pure Reinforcement Learning, allowing it to learn reasoning through trial and error without the need for labeled data or prior supervision.
   
• Multi-Head Latent Attention: It looks at data from different angles, using multiple “heads” to find patterns and then combining them for more efficient processing.
   

Edited by : Sujatha Rao