AI system Sam Altman praises DeepSeek AI as "impressive," yet expresses reservations about its efficiency claims, particularly in comparison to OpenAI's imminent dominance.
=====================================================
In a rapidly evolving landscape of artificial intelligence (AI), the Chinese startup DeepSeek has made waves with its generative AI model that reportedly surpasses OpenAI's o1 reasoning model across various categories, including math, science, and coding, at a fraction of the development cost. However, recent reports and international concerns have cast doubt on DeepSeek's cost-efficiency and reliability.
OpenAI CEO, Sam Altman, finds DeepSeek's R1 model impressive, particularly regarding its cost-effectiveness. Yet, he expresses skepticism about DeepSeek's claims of significant efficiency, stating that he does not believe the Chinese startup has figured out something significantly more efficient than OpenAI.
Performance evaluations have revealed limitations in DeepSeek's reliability and robustness. In clinical evaluations, DeepSeek-V3 showed varied performance compared to orthopedic surgeons and other AI models, with its accuracy notably lower than human experts (44%) but somewhat higher than its AI peers like ChatGPT-4o (40%) and Gemini Pro (28%).
Governments in the Czech Republic and Italy have imposed restrictions on DeepSeek, citing national security and privacy concerns. Germany has urged its removal from app stores, and a U.S. government memo in 2025 indicates ongoing scrutiny about how closely DeepSeek aligns with Chinese Communist Party narratives, raising flags about the use of state-driven content biases.
There are also suspicions that DeepSeek leveraged the distillation technique from rival companies to save on training costs. While specific details about DeepSeek’s reported spending or financial practices are not explicit, these concerns collectively undermine the company’s credibility.
OpenAI itself has faced challenges, with projections indicating it was on the brink of bankruptcy with projections to make a $5 billion loss by the end of 2024. However, a round of funding with Microsoft, NVIDIA, SoftBank, and other key stakeholders kept its business operations afloat, pushing its market capitalization to $300 billion.
In contrast, OpenAI CEO Sam Altman remains confident that his team already knows how to build and develop artificial general intelligence (AGI) and is now focused on chasing down superintelligence. He finds the DeepSeek team very talented but believes that OpenAI will continue to dominate the AI landscape, promising to deliver better models.
Despite the challenges facing DeepSeek, the AI race continues, with both companies pushing the boundaries of what AI can achieve.
[1] Performance and Credibility Issues: [Citation Needed]
[2] Security, Privacy, and Political Concerns: [Citation Needed]
- Microsoft, as a key stakeholder in OpenAI, is anticipated to play a significant role in the development and updates of future Windows 11 software, aiming to maintain its PC market dominance, especially given the ongoing AI race.
- In light of the recent controversies surrounding DeepSeek's R1 model, concerns have been raised about the potential security and privacy risks associated with using their technology, such as on Xbox gaming consoles, due to the link with the Chinese Communist Party and alleged national security breaches.
- As open-source software becomes increasingly popular in the ever-expanding technology sphere, there are debates on whether programs like Codex – AI-powered code writers by Microsoft – help democratize software development or potentially create unfair competition for human developers.
- Following the success of OpenAI in breakthroughs in artificial intelligence, there are discussions revolving around the implications of different countries' investments in AI technology on long-term technological aegemony, with China's DeepSeek serve as an example for this debate.
