Operators strengthen AI services and usher in an era of large-scale "blooming"

The era of large model "flowers" is here! Operators deeply develop AI services

OpenAI announced the official launch of GPT Store to help expand the application scenarios of large artificial intelligence models. Wu Hequan pointed out that large models have become a new trend of the Internet and will profoundly change the way of human-computer interaction.

The large model industry is booming. As an operator with cloud-network integration advantages, it is actively exploring large model services.

Large model becomes a breakthrough point for operators’ business

By opening the artificial intelligence chat window, we can use the universal assistant to achieve functions such as knowledge answering, article polishing, and English translation. In addition, we can also enter the live broadcast room where the virtual digital lecturer "Xiaolu" will teach around the clock and provide personalized feedback and suggestions based on the students' learning data. The emergence of large models marks that general artificial intelligence is gradually becoming a reality. These intensively updated large models are being combined with real-life scenarios, showing great potential and advantages on the application side. This development is a turning point in a new outbreak of artificial intelligence, allowing us to see the emergence of general artificial intelligence in practice.

Looking back at the development of the large model industry over the past year, not only the large models launched by technology companies such as Alibaba Cloud, Baidu, iFlytek, etc., but also operators can be seen. What is the significance of this move by operators?

Operators' layout of large model domains not only caters to the general trend of change in the information age, but is also an inevitable choice from the perspective of commercial value.

AI large models add strong impetus to the digital and intelligent transformation of telecommunications networks, and are very beneficial from the perspective of application value. Modern telecommunications networks continue to increase in size and complexity, resulting in more complex network deployment and operations. Therefore, the digital transformation of the network is imperative. By applying large AI models, network resources can be better managed and optimized, and network efficiency and performance can be improved. This will bring huge changes and advancements to telecommunications networks.

From a business perspective, although the "second curve business" of the three major operators has become the main driving force for performance growth, traditional businesses such as mobile phone traffic and fixed-line broadband are still the main source of operators' profits, while emerging Businesses such as cloud computing and the Internet of Things are still in their early stages of development. According to predictions from Big Model House, the global artificial intelligence large model market will reach US$21 billion by 2023 and reach US$109.5 billion in 2028. In an era when artificial intelligence is becoming popular, making full use of the "traffic" of large models will become a breakthrough for operators' business.

Huge potential for industrial applications

Wu Hequan, an academician of the Chinese Academy of Engineering, said that basic large models have strong general knowledge capabilities, but chatting is difficult to achieve rigid needs. To reflect the value of large models, it must be implemented in industry applications. Among them, the application of large models in the field of industrial manufacturing shows great potential.

New industrialization is a systematic project, an integrated development of informatization and industrialization, and a deeply integrated development of the real economy and the digital economy. Industry insiders generally believe that on the road to new industrialization, it is necessary to strengthen demand and scenario traction and liberate the creativity of employees, and large models will bring a new paradigm of intelligent production to enterprises. Take mining operations as an example. During the coal mining process, improper operations by workers, such as insufficient depth and wrong angles, will cause safety hazards and even accidents. The application of AI large models allows workers to complete standardized verification of pressure relief projects while sitting in the office, achieving intelligent management and intelligent production efforts, thereby improving the safety and production efficiency of the mining industry.

Currently, operators’ large models mainly focus on predictive maintenance, industrial quality inspection, production process intelligence, digital production assistants, intelligent supply chain and logistics, digital twins, data security and compliance, and sales recommendations through vertical industries. and automation, intelligent customer service and other empowerment to achieve industrial digital transformation and promote the rapid development of new industrialization.

China Telecom’s Galaxy AI platform has accumulated thousands of visual recognition algorithms and has been widely used in industrial scenarios such as textile quality inspection and logistics sorting. China Mobile has developed over 370 AI capabilities in various fields such as machine vision, speech, natural language processing, and network intelligence. China Unicom has also self-developed more than 60 core algorithms in machine vision, language recognition, natural language processing, and knowledge understanding.

The surge of large model open source

Looking around the world, there are two routes to building an artificial intelligence ecosystem: open source and closed source. Due to its uncertainty, large models form a basic pattern in which the open source model and the pre-research model are promoted simultaneously.

Training large models requires a lot of computing power and high development costs, which to a certain extent prevents small and medium-sized enterprises from developing large models. At this time, large model open source can reduce the cost of enterprise deployment of large models, and can improve the application of large models to a certain extent.

For a long time, open source has become the first mainstream model in the Internet era. In 2017, China released the "New Generation Artificial Intelligence Development Plan", which included open source and openness as basic principles in the plan. Wu Zhaohui, Vice Minister of the Ministry of Science and Technology, said at the 2023 Zhongguancun Forum that China adheres to open source collaboration, strengthens continuous innovation of large model technology, collaboratively solves common problems such as transparency and stability, further promotes the open sharing of computing resources and digital resources, and accelerates the formation of large models industrial ecology.

Based on long-term considerations, Huang Tiejun, director of Beijing Zhiyuan Artificial Intelligence Research Institute, once said that the technology itself is constantly evolving iteratively and cannot be closed to a few companies. Third-party participants gather into a torrent and may become a technology in the big model era in the future. base.

It is worth noting that just recently, China Telecom announced that it will fully open source its self-developed Xingchen AI large-scale model to better assist the intelligent upgrade of customer business. At present, China Telecom has announced that the Xingchen AI large model will be open sourced to the whole society through Github, Gitee and Huggingface platforms, and will disclose its underlying code, algorithm logic and various pre-made basic large models, development modules, training tools and other core product capabilities.

(Picture source: Photo Network)

END

0

To submit articles to "Communication Information News", please write to: txxxb2001@163.com,

Once the manuscript is published, it will be determined based on the quality of the article

Provide remuneration of 200 yuan to 500 yuan per thousand words.

Other cooperation, suggestions, news leads,

Welcome to contact us in the background of WeChat official account.

Bad information reporting hotline: 0591-83365173.

The above is the detailed content of Operators strengthen AI services and usher in an era of large-scale "blooming". For more information, please follow other related articles on the PHP Chinese website!