DeepSeek Challenges AI Assumptions, But Big Tech Could Still Benefit

The Chinese startup’s innovative AI model raises new questions about the future of AI investments, potentially playing to Big Tech’s advantage.

On Monday, OpenAI CEO Sam Altman posted a message on X that acknowledged a significant shift in the AI landscape, confirming what many in the tech world had not anticipated—DeepSeek, a Chinese AI startup, has shaken up the industry.

“DeepSeek’s R1 is an impressive model, especially for the price,” Altman wrote, giving credit to a surprising new competitor. "We will obviously deliver much better models, but it’s invigorating to have a new competitor. We’ll pull up some releases."

DeepSeek, which surged to prominence last week, launched a groundbreaking AI assistant that quickly became a top app on the Apple App Store. Its success has sparked reactions from major tech figures, including former President Donald Trump, and has raised eyebrows across Silicon Valley.

The company’s rise calls into question the core assumptions that Big Tech has made about AI’s future—a future dominated by massive investments in hardware and data centers. The traditional approach in Silicon Valley has focused on refining AI models using expensive, energy-intensive computing power, typically relying on powerful chips and enormous data centers. This approach has led to projections of over $1 trillion in AI investment, with industry leaders like Mark Zuckerberg and Elon Musk committing billions toward advancing AI capabilities.

However, DeepSeek has taken a radically different approach. The Chinese startup, operating with a small team of talented engineers recruited directly from universities, has managed to create a competitive AI model without relying on high-end chips or energy-guzzling training methods. This efficiency challenges the current thinking about what it takes to build and operate AI models, particularly the assumptions about cost, energy consumption, and the need for cutting-edge hardware.

DeepSeek’s emergence raises several critical questions for Big Tech: What if AI models aren’t as difficult or expensive to build as previously believed? What if AI technology can be deployed on more accessible platforms like smartphones? And how has China managed to sidestep export controls that were intended to curb its AI development?

While the shockwaves of DeepSeek’s arrival have rattled some sectors of the tech industry, not everyone is convinced that it marks the end of the road for Big Tech. Some experts argue that the commoditization of AI could actually benefit major players like Microsoft, Google, and Nvidia in the long run.

Ben Thompson, a business and technology analyst, explained in his blog, Stratechery, that while DeepSeek’s model may have disrupted the initial AI landscape, the wider adoption of cheaper AI technology could prove advantageous for Big Tech. “A world where Microsoft gets to provide inference to its customers for a fraction of the cost means they can spend less on data centers and GPUs—or, just as likely, see dramatically higher usage,” Thompson wrote.

Even former President Trump echoed this sentiment, pointing out that “instead of spending billions and billions, you’ll spend less, and hopefully, come up with the same solution.”

Christopher Ackerman, an independent AI researcher and former Google employee, also weighed in, suggesting that while the competition may intensify for startups like OpenAI and Anthropic, the need for chips will persist, providing a safeguard for companies like Nvidia. DeepSeek itself used less advanced Nvidia chips in its own model training, a fact that further complicates the narrative.

Despite the initial market shock, analysts like Dan Ives of Wedbush Securities believe that the attention given to DeepSeek’s disruption is overblown. He argues that the continued innovation in AI will lead to lower model costs, a trend that will ultimately benefit computing power and AI use cases in the long run.

As DeepSeek continues to challenge established norms, it’s clear that the AI landscape is evolving. Whether this disruption will be a lasting challenge for Big Tech or an opportunity for innovation remains to be seen—but for now, the industry’s giants are not backing down.