Although ChatGPT is not innovative in terms of underlying technology, its social impact has far exceeded expectations. This chat robot developed by OpenAI, an American artificial intelligence company, has become the fastest growing consumer application in the world since its launch in November 2022. It has always been one of the important goals in the field of artificial intelligence to let machines and real people talk freely. Behind the explosion of ChatGPT is the development of deep learning technology for ten years. Not long ago, at the Baidu Create AI developer conference, Wang Haifeng, director of the National Engineering Research Center for Deep Learning Technology and Application and chief technology officer of Baidu, said that the current large-scale AI production has been formed, and deep learning has gradually matured in technology, ecology, industry and other dimensions. The technological innovation and industrial development of AI has entered the "deep learning+" stage. Deep learning allows AI application fields to further understand "deep learning+", first of all, what is deep learning. Compared with the traditional machine learning algorithm, the deep learning based on neural network algorithm is "deep". Although the application of traditional machine learning algorithms in fingerprint recognition, face detection and other fields has basically reached the commercial requirements, it is difficult to "go further" until the emergence of deep learning algorithms. Deep learning belongs to unsupervised learning, which can be automatically completed without manual sample labeling. It should be pointed out that deep learning is very dependent on hardware facilities, because it requires too much computation and requires a lot of time and data to train. A technology can extend its tentacles to all walks of life, depending on its underlying universality. "Deep learning has a strong universality, showing the characteristics of standardized, automated and modular industrial mass production." Wang Haifeng pointed out that large-scale AI mass production has been formed from the perspective of broader support value. The technological innovation and industrial development of AI have entered the stage of "deep learning+". Deep learning promotes the accelerated development of various industries. Deep learning enables machines to learn from massive data and large-scale knowledge at the same time, with better results and higher efficiency. For example, the Wenxin industry-level knowledge enhancement model developed by Baidu has deep semantic understanding and generation capabilities across modes and languages, which can be applied to Internet products such as search, information flow, smart speakers, and enable manufacturing, energy, finance, communications, media and other industries through the deep learning platform. Chips, deep learning frameworks, models and applications form a sound ecosystem of deep learning, enabling application needs and feedback to be transmitted to all links of deep learning technology, and each link continues to be optimized iteratively to accelerate AI technology innovation and industrial development. In addition, all parties involved in production, education, research and application in ecology are also working together to cultivate AI talents. All walks of life apply in-depth learning technology to reduce costs and increase efficiency, innovate products and businesses, accelerate the process of industrial intelligence, and strive to achieve high-quality growth. China's industrial system has a complete range and a large volume. The innovative application of artificial intelligence driven by deep learning helps to form a virtuous circle of industry, promote breakthroughs in underlying technologies, and accelerate the upgrading of modern industrial system. For example, intelligence