DeepSeek, a relatively obscure Chinese AI startup, is shaking the foundations of the global AI market with its latest announcement: it has developed AI models nearly on par with the best from Silicon Valley, but at less than half the cost. Priced at just $6 million, this has raised eyebrows across Wall Street, especially among competitors like AMD and Nvidia, who have been investing billions to stay at the cutting edge of AI technology.
DeepSeek’s emergence was timely, coinciding with AMD’s scheduled release of its fourth-quarter earnings report, expected on February 6, 2024. The company has struggled recently, with its stock plummeting 34% over the past year, compared to Nvidia’s meteoric rise of 68%. Amidst this backdrop, DeepSeek’s news sparked significant investor concern about sustainability and the inflated valuations of AI stocks.
“We need to deploy models with 95%-99% accuracy,” noted Rohan Malhotra, CEO of Roadzen, highlighting DeepSeek’s high level of performance. Malhotra’s company has been employing various AI technologies to streamline insurance claims processing, and DeepSeek's technology has proven to be game-changing.
“DeepSeek just lowered our costs by 90% so we can help more people,” stated Neal K. Shah, CEO of CareYaya, who has leveraged the startup’s technology to combat health insurance claims denials. Such statements reveal just how transformative DeepSeek’s innovations are proving to be, particularly for businesses seeking greater efficiency.
Roadzen alone processed over 607,000 claims last quarter, where using DeepSeek-R1 resulted in 50% lower costs compared to using OpenAI’s models. This is significant for insurers who often operate on razor-thin margins and need tools to optimize their operations.
Nonetheless, the ripple effects of DeepSeek’s advancements have reached far beyond individual companies. Following the announcement, AMD’s stock reacted negatively, continuing its downtrend as pressure from President Donald Trump’s imposed tariffs on China looms overhead. “Businesses and consumers might be deterred by the increased prices, affecting chip manufacturers like AMD,” analysts warned.
Adding to the concerns, Ryan Taylor, the Chief Revenue Officer of Palantir, stated, “We would strongly discourage it (DeepSeek) and don’t think any customer in the U.S. government will be able to use it,” referencing the national security risks perceived with utilizing DeepSeek's technology. The rapid advancement of AI by non-U.S. firms has caused trepidation among government and defense contractors, leading to increased scrutiny.
DeepSeek’s role as an open-source innovator is likely to inspire other companies to rethink their AI strategies entirely. Arli Charles Mujkic, CEO of Swedish firm Ooda AI, embraced the startup's technology almost immediately, integrating it the same day it launched. “This is the kind of paradigm shift that's happening now,” he remarked, indicating the potential broader industry trend of increasing open-source development.
The cost of AI is significantly decreasing, and the potential fallout might reshape the competitive dynamics among major tech players, with many now revisiting the cost versus output efficiency of their proprietary systems.
Despite the cloud of uncertainty surrounding tariffs, AMD is expected to report earnings of $1.09 per share on revenues of $7.5 billion, with its data center business projected to generate significant growth. Nonetheless, questions remain about whether the company can compete effectively against the rise of DeepSeek and other low-cost AI technologies.
Investors, who initially expressed panic as they dumped shares following the news, might soon be reminded—the AI race is far from over. Each twist and turn could create unique opportunities or setbacks as tech giants fight to maintain their foothold amid fast-evolving conditions.
Overall, DeepSeek’s low-cost AI capabilities present compelling advantages for businesses, potentially challenging the established incumbents to rethink their long-term viability and strategies. Whether AMD can recover from this shockwave and adapt to the changing market dynamics remains to be seen, but it is clear: innovation, cost, and performance are at the forefront of this AI revolution.