On January 20th, a small team of Chinese developers released what many experts are calling one of the most advanced AI language models thus far: DeepSeek-R1. Previously, the industry was dominated by the likes of OpenAi (GPT) and Anthropic (Claude), but the improbable emergence of this new Chinese startup has raised questions regarding their perceived superiority.

However, the main shock stems not from the computation capabilities of Deepseek, but instead, the cost efficiency it exhibits in comparison to rival AI behemoths. At a meager budget of only $6 million, the cost to train its AI model is a mere drop in the ocean when compared to the billions of dollars pumped into Silicon valley AI endeavours. Furthermore, during the developmental phase, the company lacked sufficient access to the most advanced computational chips, primarily due to a U.S. imposed ban on the sale of said technology to China. 

Thus, DeepSeek’s achievements have undercut the belief that bigger budgets and top-tier chips are the only ways of advancing AI, a prospect which has created uncertainty about the future of high-performance chips. “DeepSeek has proven that cutting-edge AI models can be developed with limited computing resources,” says Wei Sun, principal AI analyst at Counterpoint Research. Because of this, investors were in a frenzy to pull their money out of stocks as the perceived value of companies associated with AI plummeted.

This frantic selloff wiped over 1 trillion dollars from the stock market in total and caused the NASDAQ to tumble over 3%. Individually, Chip-maker NVIDIA took the largest hit as its market value fell a record-breaking amount: nearly 600 billion dollars and more than GDP of most countries. Such a dramatic and volatile swing in a previously stable part of the economy alludes to a broader overvaluation of the AI industry and has impelled investors to reassess the value of the tech market as a whole. 

Nevertheless, despite the detriments of Deepseek’s unprecedented arrival at the forefront of artificial intelligence development, the future for innovation is bright as an “AI-arms race” looms. Trump said on Monday DeepSeek’s ability to train AI more cheaply is a “wake up call” for tech industries. Sam Altman has also already responded, announcing that OpenAi would accelerate the release of “better models.” “In China, DeepSeek’s advances are being celebrated as a testament to the country’s growing technological prowess and self-reliance,” says Marina Zhang, an associate professor at the University of Technology Sydney.