China Challenges Trump’s AI Policies with Fast and Powerful DeepSeek

It’s called DeepSeek R1, and it’s making big news, especially on Wall Street.

What is DeepSeek R1?

DeepSeek R1 is an AI system created by DeepSeek, a new Chinese company that’s only been around for a year. Despite its young age, it has achieved a big win that famous tech investor Marc Andreessen calls “AI’s Sputnik moment.” This AI is almost as smart as big-name systems like OpenAI’s GPT-4, Meta’s Llama, and Google’s Gemini, but it costs far less to run.

DeepSeek said it spent only $5.6 million to power its core AI system. Compare that to the hundreds of millions or even billions of dollars US companies spend. What’s even more surprising is that China has limited access to the best AI chips due to US restrictions. Yet, DeepSeek managed to build its system using less advanced chips.

Who Created DeepSeek?

DeepSeek was started in late 2023 by Liang Wenfeng, a hedge fund manager in China. His company is one of many startups trying to ride the AI wave, getting money from investors and working on groundbreaking projects.

Liang is now being compared to Sam Altman (the head of OpenAI) in China. His hedge fund, High-Flyer, focuses on building and funding AI technology.

Before DeepSeek R1, the company released other AI tools, but they didn’t get much attention. One earlier version, V3, sparked debates because it avoided sensitive topics related to the Chinese government. But R1, launched late last year, has taken everyone by surprise.

What’s more, DeepSeek R1 is open-source. This means other companies can use it, test it, and improve it.

Why Is DeepSeek a Big Deal?

AI systems usually need huge amounts of money and energy to work. Some American tech companies are even buying nuclear power plants to keep their AI tools running. For example: Meta plans to spend $65 billion on AI this year.

Sam Altman from OpenAI said the AI industry will need trillions of dollars in investment for the chips and electricity needed to power these systems. But DeepSeek has shown that a powerful AI can be built for far less money and with cheaper chips. This changes how people think about the cost of AI and how much is truly needed to compete.

Tech investors and experts are stunned. Andreessen called DeepSeek one of the most impressive breakthroughs ever. If powerful AI can be built at such a low cost, it could open new opportunities—and even risks—for the whole world.

Impact on America

The US has tried to stay ahead in AI by restricting China’s access to advanced AI chips, claiming it’s a matter of national security. Former President Joe Biden even increased restrictions on chip exports before leaving office.

But DeepSeek’s success suggests these restrictions might not work as planned. Despite the chip bans, China’s AI capabilities seem to be catching up quickly.

This news sent shockwaves through US financial markets: Nvidia (a major AI chip maker) saw its stock drop by 12% before markets opened. Other big companies like Meta, Google’s parent company Alphabet, and Oracle also saw their stock prices fall.

Is DeepSeek’s Success Overhyped?

Some experts are skeptical about how DeepSeek kept its costs so low. While the company revealed its operating costs, it hasn’t shared how much it spent on training the model, which can be very expensive. Even so, the cost was likely far lower than what US companies spend.

It’s also too soon to say if this one success will hurt America’s long-term leadership in AI. US companies still lead in terms of talent, research, and infrastructure. Plus, DeepSeek R1 is mainly a consumer-focused tool (like ChatGPT). It hasn’t yet proven it can handle larger, more complex tasks that industries need.

For now, DeepSeek’s breakthrough is a wake-up call to the AI industry—but the race isn’t over yet.