Bahattin Gonultas
January 27, 2025•Update: January 27, 2025
BERLIN
China’s newly unveiled AI chatbot, DeepSeek, has raised alarms among Western tech giants, offering a more efficient and cost-effective alternative to OpenAI’s ChatGPT.
Supported by the Chinese hedge fund High-Flyer, DeepSeek launched its DeepSeek-R1 large language model (LLM) on Jan. 20. Unlike ChatGPT’s subscription-based and closed-source platform, priced at $200 per month, DeepSeek-R1 is entirely open-source and free, allowing users to access, compile, and operate it on native hardware without limitations.
Experts report that DeepSeek-R1 surpasses ChatGPT and other leading models, including Google’s, in key performance benchmarks. Following its release, the app rapidly climbed into the top 10 rankings on mobile app stores.
The model was developed and trained in just two months using only 2,000 Nvidia chips at a cost of $5.6 million. This contrasts sharply with the significantly higher expenses of companies like OpenAI, Meta, and Google, which spend roughly 10 times as much on proprietary models.
Jim Fan, a research scientist at Nvidia, described the R1 model as “the biggest dark horse” among the open-source Large language models (LLMs), underscoring its potential to transform the global AI industry.
US chip restrictions drive innovation
Analysts attribute DeepSeek’s achievements to US-imposed chip restrictions on China, which pushed the company to maximize efficiency. This optimization, coupled with its open-source nature, is reshaping the competitive landscape and challenging the dominance of Western tech firms.
DeepSeek’s introduction has also sparked a price war in China, pressuring major companies like Alibaba, ByteDance, and Baidu to respond with similarly cost-effective models.
Global implications
Experts caution that the rise of DeepSeek could significantly impact the revenues of companies like Google, OpenAI, and Nvidia, as affordable AI models reduce the demand for expensive proprietary systems. Some even predict a potential repeat of the “dot-com bubble” crisis that shook US tech firms in 2000.
Notable figures, including Marc Andreessen, a venture capitalist, hailed DeepSeek-R1 as “one of the most amazing and impressive breakthroughs” and “a profound gift to the world.”
Meanwhile, Microsoft CEO Satya Nadella, who plans to invest around $80 billion in AI this year, speaking at the World Economic Forum, acknowledged DeepSeek’s impressive performance, emphasizing the importance of recognizing China’s growing role in AI development.
As competition intensifies, DeepSeek’s disruptive capabilities signal a transformative era for the global AI sector.
*Writing by Emir Yildirim
