DeepSeek, which launched in 2023 and is based in China, has really shaken things up in the AI market since 2025. DeepSeek, founded by Liang Wenfeng, is not only a major contender for AI dominance but has also spurred conversations about the future of AI between Western and Eastern companies. If you’re even the least bit interested in how AI will develop, I suggest you check out my article. You’ll find out a lot more about DeepSeek there.
Liang Wenfeng is an electronics student turned successful businessman. In 1985, Wenfeng was born in Guangdong, China. While he was at Zhejiang University, his attention drifted toward artificial intelligence and automation. As a consequence, he co-founded the highly successful High-Flyer Quant hedge fund in 2015. Fast forward a few years to 2023, when he co-founded DeepSeek, an artificial intelligence company with the mission of making artificial general intelligence (AGI) accessible to everyone, everywhere.
DeepSeek-R1, a cool AI, is making headlines in the field. It has drawn attention for its impressive abilities in reasoning in complex domains such as mathematics and computer programming. During benchmark tests, DeepSeek-R1 has demonstrated that it can hang with OpenAI’s ChatGPT. Plus, DeepSeek-R1 was developed for much less (around $6 million) than the cost of ChatGPT’s maker, GPT-4 (around $100 million).
The answer lies in the roughly 2,000 Nvidia H800 chips DeepSeek employs. This simple approach is pretty impressive, given the supply chain crunch the world is dealing with.
DeepSeek-R1’s ability to function like a personal assistant, free of charge, has the tech industry on edge. The recent top spot on Apple’s App Store came after crushing the competition. The free AI has raised questions: are big bucks and powerful computers essential for the next generation of technology?
A distinguishing feature of DeepSeek is its commitment to openness. Unlike many Western AI firms that maintain proprietary research, DeepSeek has made its code publicly available. This transparency allows researchers and developers worldwide to access, modify, and build upon their models, potentially accelerating innovation across the AI community. However, this openness also raises concerns among U.S. companies and policymakers about maintaining a competitive edge and addressing national security implications.
Industry leaders have reacted to DeepSeek’s achievements in mixed ways. Marc Andreessen, a venture capitalist, called the DeepSeek-R1 “one of the most impressive and amazing breakthroughs I’ve ever seen,” emphasizing the model’s importance in the field of artificial intelligence. On the other hand, some leaders, such as Alexandr Wang, CEO of Scale AI, view DeepSeek’s emergence as a “wake-up call for America.” Wang calls attention to the concerns about the shift in technological leadership.
Despite its successes, DeepSeek faces scrutiny over certain aspects of its model. Critics point out that DeepSeek-R1 aligns its responses with Chinese government perspectives, particularly on sensitive political topics, leading to concerns about censorship and bias. Additionally, while the model excels in various tasks, some analysts note that it lacks certain capabilities present in more extensively trained models, suggesting areas for further improvement.
DeepSeek’s rapid ascent underscores the evolving dynamics of the global AI industry. By delivering a high-performing model at a reduced cost and embracing an open-source philosophy, DeepSeek challenges established norms and prompts a reevaluation of strategies within the AI community. As the company continues to develop and refine its models, its impact on both the market and technological discourse will be closely monitored by stakeholders worldwide.