【Abstract】The release of ChatGPT triggered a generative AI boom, representing the miracle of general artificial intelligence. At this moment, the information industry ecosystem is very likely to be restructured. Domestic industry circles have strengthened research in the field of intelligent computing represented by ChatGPT. Operators, as the main force in the construction of computing network infrastructure, will usher in new opportunities for the development of intelligent computing. . It analyzed in detail the capabilities, development and application prospects of generative AI, and considered the technical elements behind generative AI, the demands for computing network resources, the impact on the communications industry, and the role of operators in the wave of generative AI development. Opportunities and challenges faced, and finally the operator's positioning and response strategies were discussed.
【关键词】ChatGPT ; Generative AI ; Communication work ; Wisdom calculation
0 Preface
At the end of November 2022, OpenAI released a dialogue program (chat generative pre-trained transformer, ChatGPT) based on the generative pre-trained transformation model for artificial intelligence dialogue applications. It demonstrates intent understanding capabilities, language fluency, and continuous dialogue capabilities. It swept away the image of traditional natural language processing (NLP) with weak application capabilities and became another explosive event after AlphaGo defeated Lee Sedol in 2016, setting off a craze for large models and generative AI in the global AI industry. As of April 2023, more than ten similar large models have been released at home and abroad.
Practitioners in the communications industry are also thinking about how to treat generative AI based on large models, and how telecom operators' computing network infrastructure needs to respond. Based on the analysis of generative AI development trends, key technical elements, and future application scenarios, this article will propose the key requirements of generative AI for computing power, networks, and data centers, and then propose suggestions for telecom operators to respond to this round of AI development boom.
Since generative AI is in a period of rapid development, events that subvert human cognition continue to occur. Based on