Recently, Meta engineers expressed concerns about Chinese AI company DeepSeek and its model R1 on an anonymous platform. R1 has caused industry shock with its low training costs (only $5.5 million) and performance comparable to OpenAI's o1 model, especially when it is open source far exceeding OpenAI. This incident highlights China's rapid rise in the field of AI and poses unprecedented challenges to American tech giants. The annual salary of Meta executives even exceeds the training cost of DeepSeek V3, highlighting this pressure.
In contrast, Meta executives often exceed the training costs of the entire DeepSeek V3, which makes Meta management embarrassed.
According to internal information, the emergence of DeepSeek V3 has caused Meta to feel a lot of pressure last year. Meta engineers are stepping up their time to analyze DeepSeek's technology, hoping to replicate the key technologies as soon as possible. In the past, AI researchers around the world were chasing the big model of the United States, but now the situation has changed, and American engineers have begun to reverse engineer China's AI technology.
DeepSeek not only surpassed Meta's Llama4 in benchmark tests, but also attracted industry attention in its performance in multiple fields. DeepSeek achieves extremely high inference performance through technologies such as large-scale reinforcement learning (RL) and unsupervised fine-tuning (SFT). This technological confidence has caused some American netizens to begin to reflect on China's rapid rise in the field of AI.
Points:
The training cost of the DeepSeek R1 model is only $5.5 million, and its performance is comparable to OpenAI's o1.
The executives at Meta exceed the cost of training at DeepSeek, and management is under great pressure.
DeepSeek's success has sparked panic among American tech companies and challenged the U.S. dominance in the AI field.
The emergence of DeepSeek R1 marks a major change in the global AI landscape, with its low cost, high performance and open source features that have a huge impact on existing AI giants. This is not only a competition at the technical level, but also reflects the intensification of competition in the global AI industry and the quiet changes in the balance of power. In the future, the continuous development of DeepSeek deserves continuous attention.