Tech

The Chinese AI Model ‘DeepSeek’: A New Revolution Disrupting Tech Giants

Chinese companies have been working on developing their own AI systems since the launch of “ChatGPT.” While Alibaba had previously introduced a model, it failed to achieve the desired competitive level, resulting in little global attention.

However, the emergence of the “DeepSeek” model has completely changed the landscape, causing a global stir that has impacted all Silicon Valley AI companies. Described as more powerful and efficient than models like “ChatGPT” and “Gemini,” what’s the full story behind this revolutionary model?

An Exceptional Model Changing the Game

The “DeepSeek” model is developed by a Chinese startup of the same name. Although its official app launch happened last week, its journey began weeks earlier with a trial phase open to global users. The company was founded by Chinese investor Liang Wenfang, a prominent figure in the AI sector and a co-founder of the “High Flyer” investment fund.

In April 2023, Wenfang launched the company with the goal of achieving Artificial General Intelligence (AGI), a goal pursued by all AI companies. In a short time, the app achieved immense success, topping the Apple App Store as the most downloaded app in the U.S., surpassing popular American apps, particularly during the ongoing TikTok ban controversy.

Low Cost: The Key to Success

Currently operating under version “R1,” DeepSeek competes with “ChatGPT-01” at a fraction of the cost. DeepSeek costs users just $0.50 per month compared to $20 for ChatGPT. The company attributes this affordability to the relatively low development and operational costs of $5.5 million, compared to the hundreds of millions spent by companies like OpenAI and Google.

Breaking American Dominance

Despite U.S. sanctions aimed at limiting Chinese companies’ access to advanced technologies, DeepSeek managed to develop a model capable of logical reasoning without relying on Nvidia’s U.S.-made chips. Estimates suggest that the company may have developed its own specialized chips or employed innovative solutions that bypass the need for these components, putting companies like OpenAI and Meta in a tough position, especially given the model’s significantly lower development costs.

Challenges and Pressures

The model’s rapid success attracted millions of users in a short time, prompting the company to temporarily suspend new registrations to ensure service stability. The app also faced cyberattacks that affected its performance. Despite these challenges, the model is expected to continue thriving as it expands its services in the future.

Impact on the AI Sector

DeepSeek’s debut has sent shockwaves through the AI industry. Following its launch, Nvidia’s stock dropped by 17%, amounting to losses of $600 billion, affecting other companies like Marvell and Broadcom as well. These losses have forced American companies to reassess their strategies, with executives like Microsoft CEO Satya Nadella acknowledging that DeepSeek’s success signals a major shift in the industry.

U.S. Companies’ Response

Sam Altman, CEO of OpenAI, admitted that DeepSeek has achieved remarkable progress but emphasized that the competition remains ongoing with the development of new features. Meanwhile, Yann LeCun, Chief AI Scientist at Meta, described the success as a victory for open-source AI technologies.

A Message to the World

DeepSeek’s success sends a powerful message to global companies: innovation is not exclusive to any one player, and the race to develop AI remains open to everyone.

Back to top button