With the World Economic Forum having concluded in late January, the business world’s eyes turned from Davos to Paris. The Paris AI Action Summit took place on February 10th and brought industry leaders, CEOs, academics, and nonprofits together to discuss the future of artificial intelligence. Beyond the private sector implications of this influential summit, world leaders also attended to gauge and lead discussions about the future of artificial intelligence in a world with a growing reliance on it. While French President Emmanuel Macron and Co-chair Indian Prime Minister Narendra Modi are pushing for tech sovereignty, mainstream media reduces the AI race to a struggle between two superpowers: China and the United States.
Many considered the United States to be in pole position in the AI race, fueled by significant private sector funding, a bustling startup ecosystem, and massive research networks between elite academic institutions. The crowning jewel in the United States AI arsenal is OpenAI, a pioneer in accessible AI models and generative AI with programs such as DALL-E, Codex, and Vanderbilt students’ favorite, ChatGPT. However, while OpenAI was the driver of the U.S.’s AI strategy, its engine was undoubtedly Nvidia, a stock market darling, that dominated AI hardware, providing the GPUs powering nearly every significant AI model, primarily through their H100 and A100 chips. The dominance of these AI supergiants led to massive valuations, with OpenAI’s aggressive fundraising round valuing the company at $340 billion and NVIDIA achieving a whopping valuation of $3.7 trillion, rivaling Apple in value.
However, while NVIDIA and OpenAI’s dominance seemed inevitable in 2025, Chinese startup DeepSeek turned heads and markets with the rollout of the DeepSeek R1, rivaling OpenAI’s most dominant model while costing far less to develop. DeepSeek’s innovation lay in its ability to reduce data processing costs while relying on lower-quality hardware, challenging the assumption that only high-end GPUs could train top-tier models. Led by Liang Wenfeng and backed by his $8 billion hedge fund, High-Flyer, the team shocked the industry by optimizing model training efficiency despite having access to inferior chips compared to those in the U.S.
The revelation of a new emerging AI superpower forced NVIDIA’s valuation to decline a whopping $560 billion, marking the most significant one-day loss in stock market history. This $560 billion collapse dwarfs previous market plunges, surpassing Meta’s historic $230 billion drop in 2022. DeepSeek’s seemingly overnight success has been a warning to Silicon Valley, with even the American government taking notice. Last week, President Trump called DeepSeek’s arrival a “wake-up call for our industries to be laser-focused on competing to win.”
DeepSeek is a sign of China’s rising AI capabilities, as massive companies such as Alibaba, Bytedance, and Tencent increase their investments in generative AI technology. Beyond the economic motives for increasing AI efficacy and efficiency, Harvard Professor David Yang indicates China’s autocratic regime would benefit from AI development for data collection and predictive powers. However, China’s autocratic nature may also pose challenges to DeepSeek, as countries like Italy and Australia have already banned DeepSeek and launched investigations into privacy concerns.
With the emergence of DeepSeek as a formidable opponent to U.S. dominance, an AI showdown in Paris was inevitable between the two global frontrunners. However, other countries are not to be overlooked in the AI sweepstakes. Under Prime Minister Modi, India has launched INDIAai, an initiative to “build a comprehensive ecosystem, democratizing computing access, enhancing data quality, developing indigenous AI capabilities, attracting top AI talent, enabling industry collaboration, providing startup risk capital, ensuring socially impactful AI projects, and promoting ethical AI.” India’s large workforce dedicated to information technology and AI projects is expected to contribute approximately $500 billion to the economy in 2025 alone. With increased government support, a bustling startup ecosystem, and DeepSeek illuminating more cost-effective methods to train models, India hopes to position itself as an emerging AI superpower.
Europe is another hotbed for emerging AI capabilities. While French President Emmanuel Macron said Europe is “not in the race today,” the French government has invested over 109 billion euros into its AI policy. European scholars, such as Cambridge professor Neil Lawrence, suggest DeepSeek’s emergence is a promising sign for European competitiveness due to the country’s strong AI talent pool in academia and business. Utilizing DeepSeek’s open-source nature, European companies can train their models in a similar form and expand their technologies across Europe.
However, what is at stake in the AI race? The AI journey will be ubiquitous in our daily lives, with applications ranging and encompassing everything from medicine to business to defense. As AI gets more efficient, the demand for improved models will increase, a concept known to economists as Jevons paradox (a fact already made clear through the predictions of Microsoft CEO Satya Nadalla). The far-reaching implications of the AI race will determine geopolitical power, economic growth, and technological advancement for years to come.
U.S.-based companies are eager for the challenge, as both NVIDIA and OpenAI have congratulated DeepSeek for their AI innovations, and promised to implement innovative models to rise to this new competition. With serious competition on the horizon, both companies are experimenting with “test-time scaling,” a process in which AI models spend more time and computational power to “reason” out prompts, leading to more advanced outcomes. The implications of “test-time scaling” could lead to the development of more advanced models that can push the current limits of generative AI.
With the AI revolution in full swing with no sign of slowing down, there is sincere optimism the field will increase the standard of living worldwide. However, as the massive players enter the AI sweepstakes from the private and public sectors alike, there are many fears of misuse. Researchers in China have recently discovered that two of the most widely used large language models (LLMs) can replicate themselves without human assistance, presenting countless opportunities alongside the chilling ramifications of humans losing control of their Frankenstein monster. The United Kingdom’s AI safety report cosigned by 96 AI specialists found that AI’s rapid growth will improve access and quality of healthcare, streamline supply chains, and drive economic growth but also pose significant cybersecurity risks.
In a world that is rapidly being shaped by artificial intelligence, the nations and leaders who harness AI’s potential while mitigating its risks will shape the geopolitical order, command economic supremacy, and define the next technological era. These emerging developments discussed in Paris may shape our generation’s version of the space race.
