On the 20th of January 2025, the same day that Donald Trump was sworn in for his second tenor as U.S president and a day that he declared as the beginning of the “golden age of America”, a little-known Chinese IT start-up released its revolutionary DeepSeek R1 free open source AI App into the market largely unnoticed but within a week it was sending shockwaves across the industry that wiped off over one trillion dollars in the market valuation of U.S tech and energy stocks.
Just a week after its release, the DeepSeek R1 app had equalled the number of downloads that Chat GPT, the leading AI technology provider, had, with users testifying that it had capabilities equal to or even better than those of Chat GPT. But the real clincher was that it is a free open-source App, which made it attractive and more accessible to many more users.
While DeepSeek was celebrating its unprecedented success the stock market took notice and several Technology and Energy company stocks took a heavy hit. NVIDIA lost $600B+ in market cap, and nuclear energy giants like Vistra (-30%) and Constellation (-20%) are reeling.
While Vistra Corp (VST) dropped 28.3% its biggest single-day loss in years, Constellation Energy (CEG) went down 20.8%, as investors rethink nuclear energy’s role in the AI boom. Veritiv Holdings (VRT) Shares also tumbled nearly 30%, as demand for power-hungry data centre infrastructure became uncertain.
The losses cumulated to become the biggest US stock market loss in history just because a small Chinese AI start-up rewrote the rules of AI and the energy demand associated with powering its infrastructure.
DeepSeek R1 has done what Silicon Valley said was impossible, with a total cost of just $5.7M compared to the billions of dollars spent by US counterparts and it’s 27 times cheaper to run. A comparative cost analysis reveals that while OpenAI costs $100+ per million tokens, DeepSeek costs less than $4 per million tokens.
DeepSeek has also made the need for massive energy associated with AI infrastructure redundant with its low-cost model. For years, we thought the future of AI would mean skyrocketing electricity demand to power thousands of GPUs which will fuel the need for new energy sources (like nuclear) and scaling of infrastructure.
But DeepSeek just flipped the script with leaner models translating to lower energy demand while fewer GPUs are required to build world-class AI infrastructure. While there is a cut in costs, it also leads to a drop in demand for energy and ultimately stock prices for nuclear energy companies.
This isn’t just a cost revolution – it’s a paradigm shift.
DeepSeek’s free open-source AI technology is sending shockwaves through the AI industry and beyond. This Chinese start-up’s innovative approach is challenging the dominance of US tech giants and reshaping the global AI landscape.
The implications of DeepSeek’s technology are far-reaching. By making its model open-source, DeepSeek is democratizing access to AI technology, enabling a wider range of developers, researchers, and organizations to leverage its capabilities. DeepSeek’s success is also challenging the long-held dominance of US tech giants in the AI space, potentially shifting the balance of power in the global AI landscape.
The ripple effects of DeepSeek’s technology would be felt across the global economy, with potential disruptions to the semiconductor, cloud services, and data center industries while its open-source approach could foster greater collaboration and innovation in the AI community, driving progress and advancements in the field.
As the AI industry continues to evolve, one thing is clear: DeepSeek’s free open-source AI technology is a game-changer. Its implications will be felt for years to come, in shaping the future of AI and the global economy.
Oshobi, a development economist, management consultant, and author writes from Lagos, Nigeria.