Recently, with the release of DeepSeek's latest open source model, DeepSeek-R1, which has attracted warm attention at home and abroad, Baidu AI Cloud, Tencent Cloud, Alibaba Cloud, Huawei Cloud and other platforms announced the launch of DeepSeek's models. Industry insiders believe that DeepSeek's new developments reveal a new trend in the competition for big models in 2025, which is expected to trigger a wave of innovation. Companies will explore cost-effective AI development and deployment methods, driving continued progress in global AI. According to the official website of DeepSeek-R1, reinforcement learning technology is widely used in the post training stage, greatly improving the model's reasoning ability with very little labeled data. In tasks such as mathematics, code, and natural language reasoning, the evaluation performance is close to the official version of the GPT-o1 model developed by OpenAI in the United States. An Yun, Deputy Director of the Artificial Intelligence Research Institute at the Saizhi Industry Research Institute, stated in an interview with reporters that DeepSeeker R1 has achieved breakthrough technological progress through open source strategy, low-cost and efficient reasoning, and the combination of reinforcement learning and hybrid expert architecture (MoE). Open source has broken the technological monopoly of large enterprises and promoted the popularization of AI technology. Its low-cost algorithm optimization model has changed the long-standing dependence on computing power accumulation and promoted an efficiency oriented competitive landscape. DeepSeek will usher in a new stage of global large-scale model development and application Lu Feng, the director of the Beijing Frontier Future Technology Industry Development Research Institute, believes that DeepSeek's high cost-effectiveness and low training costs greatly reduce the investment, development, and operation costs of large models. Its openness and open source have lowered the technical threshold for integrated applications, providing more possibilities for the widespread implementation and application of large models in various industries. The reporter noticed that DeepSeek has attracted the attention of many domestic and foreign companies with its powerful language processing capabilities and technological advantages. In recent days, Baidu AI Cloud, Huawei Cloud, Alibaba Cloud, Tencent Cloud, 360 Digital Security Group and other platforms have announced the launch of DeepSeek's big model. In addition, on January 31st, three American tech giants, Nvidia, Amazon, and Microsoft, announced the integration of DeepSeeker R1 on the same day. For example, Tencent Cloud stated that its TI platform fully supports one click deployment of DeepSeek series models. As an enterprise level machine learning platform, TI platform also provides model service management, monitoring and operation, resource scaling and other capabilities to help enterprises and developers efficiently and stably integrate DeepSeek models into practical business. At the same time, DeepSeek's low-cost and efficient reasoning model also affects the upstream and downstream of the AI industry, and affects the capital market. Before the Spring Festival, many investment institutions have conducted research on listed companies in the fields of AI, chips, robots and other related industries. Lu Feng stated that with the optimization of AI models brought by DeepSeek, the AI computing power on local devices is expected to be improved, promoting the upgrading of smart terminal industries such as personal computers, smartphones, smart speakers, and smart watches, obtaining stronger intelligent interaction capabilities and functional upgrades, and expanding market application space. In addition, the rise of Chinese big models represented by DeepSeek is expected to drive the upstream and downstream development of artificial intelligence industry chains such as software, chips, operating systems, and cloud platforms, and promote the construction of a domestic artificial intelligence big model industry ecosystem. In An Yun's view, the future competition of large models will shift from a simple competition of computing power to the improvement of algorithm efficiency and reasoning ability, and deep optimization algorithms will become the new focus. Among them, with the rise of the open source ecosystem, more enterprises will use the open source model to attract developers and innovators. At the same time, the collaborative innovation of hardware and software will accelerate, especially the development of special AI chips and edge computing devices, which is expected to promote the whole chain collaboration of the industry. An Yun also stated that the attention to ethical and security issues will be strengthened with technological progress, ensuring the transparency of AI and data privacy protection as important directions for future development. (New Society)