Zeyi Yang, WIRED’s senior tech writer, dives into the rise of DeepSeek, a small Chinese tech firm that launched a competitive language model, R1, challenging giants like ChatGPT. He discusses how DeepSeek achieved this innovation at a fraction of the cost, shifting the narrative around AI development. The conversation also tackles the ethics of open-source AI, the growing rivalry between the US and China, and how US trade restrictions may have fueled China's advancements. Ultimately, it's a fascinating look at a rapidly changing AI landscape.
DeepSeek's R1 model illustrates that advanced AI can be developed efficiently and economically, challenging traditional investment-dependent narratives.
The emergence of DeepSeek signifies a critical shift in the U.S.-China tech competition, raising concerns about the future landscape of AI development.
Deep dives
The Emergence of DeepSeek
DeepSeek, a relatively new player in the AI field, has emerged as a significant competitor with the release of its large language model, R1. Founded by Liang Wenfeng, this company was initially an offshoot of a successful Chinese hedge fund that utilized machine learning for trading strategies. The unique aspect of DeepSeek is its ability to produce advanced AI technology at a fraction of the cost compared to established companies like OpenAI, with reports stating the budget for R1 was only around $6 million. This development challenges the narrative that large investments and extensive data centers are prerequisites for creating powerful AI models, thereby stirring significant interest and concern within the industry.
Revolutionizing AI Development Costs
DeepSeek's R1 model has disrupted conventional beliefs about AI development by demonstrating that high-quality models can be produced more efficiently and economically. The model incorporates a reasoning process that allows it to communicate its thought process while solving problems, akin to human cognitive reasoning. This transparency has garnered positive reception and showcased a contrasting approach to AI model training, one that prioritizes efficiency over sheer computational power. As a result, major companies like Nvidia, Microsoft, and Google have felt the market's seismic shift, leading to drastic stock declines as investors reassess the cost structures and competitive landscape in the AI sector.
Impact on U.S.-China AI Relations
The rise of DeepSeek is seen as a pivotal turning point in the ongoing tech competition between the U.S. and China, often referred to as an AI cold war. The advancement of DeepSeek's R1 model has raised alarms in the U.S., where it was assumed that Chinese companies lagged behind in AI technology. As American firms grapple with this unexpected shift, it ignites discussions about the future direction of AI development and the approaches companies take to gain competitive advantage. While the U.S. has focused heavily on acquiring cutting-edge hardware resources, DeepSeek's success may suggest that innovation can flourish through alternative methodologies that emphasize intelligence over investment.
A small Chinese tech company called Deepseek has upended the world of AI. Deepseek recently released a large language model that rivals ChatGP called R1 and it shot almost immediately to #1 on the app charts.
The interesting thing about it is that the company built their model really cheap and that has called into question this narrative that you need an endless supply of chips and data centres and money to develop AI.
On today’s show we’re speaking to WIRED’s senior tech writer Zeyi Yang about the deepening AI cold war between the US and China and the lingering questions about where AI is headed and what it’s good for?