As the excitement around DeepSeek continues to grow, this particular app has already made significant waves by outshining the capabilities of OpenAI’s o1 model in the fields of math, science, and coding. This has propelled it to the top of the charts as the most downloaded free app in the United States, unseating the hugely popular ChatGPT. The meteoric rise of DeepSeek has even sparked a noticeable drop in the stock prices of tech giants like Microsoft, Meta, and NVIDIA.
One remarkable contributor to DeepSeek’s success is its open-source R1 AI model. Meta’s leading AI expert, Yann LeCun, emphasizes that the open-source approach has been central to its breakthrough. Nevertheless, this openness has also thrown up some security red flags. DeepSeek recently had to limit new user registrations in response to a wave of “large-scale malicious attacks” targeting its service. Despite these challenges, current users can continue to access its features without any interruptions.
Within the tech community, many have lauded DeepSeek’s achievement of outperforming commercial AI models. Critics, however, argue that this success is somewhat mitigated by the open-source nature of the software, since its codebase can be freely accessed and modified by anyone. For reference, DeepSeek’s innovation is driven by its V3 open-source model.
This model’s development cost is estimated at around $6 million—a considerably smaller investment compared to the billions poured into flagship AI models, which often stall due to scaling limitations and the scarcity of high-quality training content.
The buzz surrounding DeepSeek arrives on the heels of a major initiative by OpenAI and SoftBank—the $500 billion Stargate Project. This ambitious venture aims to establish cutting-edge AI infrastructure across the United States. At its unveiling, former President Donald J. Trump hailed it as the most significant AI infrastructure undertaking in history, asserting that it would secure the technological future of the country.
The rise of DeepSeek, with its commitment to providing AI access for free, stands in sharp contrast to the debate about keeping advanced AI models closed-source—a stance that OpenAI CEO Sam Altman has defended as a safer path. The recent cyberattacks against DeepSeek certainly lend weight to his perspective on the security merits of proprietary AI systems.
