China News Service, Beijing, February 8th, ( China News reporter Song Yusheng ) Since announcing Open Source in January this year, DeepSeek’s popularity has remained high. Overseas, many industry insiders have been verifying the authenticity of DeepSeek overnight; domestically, the concept stocks of DeepSeek continued to soar this week, with multiple related stocks hitting the limit up.
How does the industry view the significance of DeepSeek? Where should Chinese AI go next? Chen Xiaoping, a professor at the University of Science and Technology of China and the director of the Artificial Intelligence Ethics and Governance Committee of the Chinese Association for Artificial Intelligence, recently accepted an exclusive interview with China News Finance.
The significance of DeepSeek is mainly manifested in two aspects. Firstly, it has strongly promoted the path of ‘low cost + Open Source’ artificial intelligence applications. Secondly, it marks a new milestone in the innovation of artificial intelligence engineering in our country.
Chen Xiaoping told reporters that before this, the R&D of the large model experienced a transformation from Open Source to closed source, and the cost has always been high.
“Worldwide, the high-cost + closed-source development path seems to be the only possible choice. Following this path, there is a high probability of creating a ‘giant unicorn economy,’ where a very small number of companies bear the majority of social production.”
In Chen Xiaoping’s view, in the ‘Giant Unicorn Economy’, there will indeed be situations where productivity is improved and the public enjoys more convenience. However, there will also be risks such as widespread job displacement and increased social polarization. The ‘low cost + Open Source’ path represented by DeepSeek provides a new possibility that is more suitable for the national conditions of developing countries. Moreover, while improving productivity and enabling the public to enjoy more convenience, the probability and extent of widespread job displacement and increased social polarization will be greatly reduced.
“Therefore, the ‘low-cost + Open Source’ path meets the requirements of high-quality development in our country. The emergence of DeepSeek has surprised the world, changing the global landscape of artificial intelligence development, which is a significant contribution of the Chinese artificial intelligence community to the world.”
On the other hand, DeepSeek is making the world rediscover China’s innovation capabilities.
Chen Xiaoping pointed out that in recent years, China has made rapid progress in artificial intelligence and large-scale model research, and has achieved a large number of results, but mainly focused on technological innovation. For large-scale AI systems such as large models, engineering innovation plays a bigger role than technological innovation. Engineering innovation needs to be application-oriented, organize and implement a large number of technological innovations, and combine the results of technological innovations to achieve significant performance improvements of large models.
In terms of engineering innovation in large models, some domestic companies have made active explorations and achieved good progress. DeepSeek’s engineering innovation has more distinctive characteristics.
He told reporters that the most noteworthy innovation in the DeepSeek project is the abandonment of blind and excessive pursuit of “generality”, and instead adopts a focused research and development strategy for natural language, mathematics, and programming, and has been successful. “In addition, the influence of low cost and Open Source has attracted global attention, replication, and recognition from peers around the world, a situation that has never occurred before.”
Chen Xiaoping predicts that under the impetus of DeepSeek, China’s artificial intelligence engineering innovation will usher in a new situation of diversified development and fierce competition, and then form a Chinese phalanx leading the global artificial intelligence engineering innovation, making greater contributions to the healthy development of the world’s artificial intelligence.
For the next development direction, Chen Xiaoping believes that the development of artificial intelligence in China faces three key focal points.
First, based on the national conditions and industrial characteristics of our country, we strive to achieve the large-scale landing of artificial intelligence industry applications as soon as possible. The high-quality development of our country requires the drive of artificial intelligence technology, especially in traditional manufacturing and smart agriculture, etc. It is necessary to actively develop industry and enterprise-specific AI systems based on the specific characteristics of segmented industries, meeting the actual needs of enterprises in terms of performance and cost. The “low cost + Open Source” path provides a historic opportunity for this.
Second, it is necessary to strengthen the governance of artificial intelligence, especially the proactive prevention of major public safety risks and major livelihood security risks, and to pay attention to the agile governance of ethical issues such as false/fabricated information generated by AI, infringement of user privacy, as well as long-term risk prediction and prevention.
Third, it is necessary to pay more attention to the basic research of large models and generative artificial intelligence, scientifically understand the fundamental breakthroughs and fundamental limitations of artificial intelligence, and provide a scientific understanding foundation for technological innovation, engineering innovation, and ethical governance.(完)
(Source: China News Network)
Source: Oriental Fortune Network
Author: China News Network