The "Hundred Model Battle" is becoming increasingly fierce, and vertical field application innovation becomes the key to breaking the situation

Our reporter Li YuchengIntern journalist Ding RongSince the beginning of this year, China's large model industry has flourished, and the popularity of the track has continued to rise. At the Artificial Intelligence Summit Forum of the 2023 Global Digital Economy Conference held recently, Jiang Guangzhi, Secretary of the Party Leadership Group and Director of the Economy of Beijing Municipal Bureau of Economy and Information Technology, said that more than 80 major models have been released nationwide

Our reporter Li Yucheng

Intern journalist Ding Rong

Since the beginning of this year, China's large model industry has flourished, and the popularity of the track has continued to rise. At the Artificial Intelligence Summit Forum of the 2023 Global Digital Economy Conference held recently, Jiang Guangzhi, Secretary of the Party Leadership Group and Director of the Economy of Beijing Municipal Bureau of Economy and Information Technology, said that more than 80 major models have been released nationwide.

The large model industry has entered the era of "Hundred Model Wars", and various parties are constantly debating and debating. Is there any duplication of construction in the 'Hundred Model Battle'? How to promote innovation in vertical field applications has become the focus of industry attention.

Several industry insiders interviewed by Securities Daily stated that the large model industry is one of the key challenges in future technological competition. For most enterprises, exploring application innovation based on large models for vertical scenarios, industries, and fields will be a key direction to tackle. Local success will drive the overall prosperity, innovation enthusiasm, and creative vitality of the industry.

Enterprises compete for layout

In the field of China's large model industry, companies such as Alibaba, Huawei, and Tencent have relatively early layouts. In 2019, Alibaba conducted research and development on large models and released the "Tongyi" large model series in 2022. Tencent also disclosed the research and development progress of its "hybrid" big model to the public in 2022.

Entering 2023, ChatGPT has sparked a wave of big models. Baidu released the big language model "ERNIE Bot" in March, becoming the first ChatGPT like product in China. Afterwards, multiple Chinese enterprises launched large-scale models to make their debut.

According to a report by Securities Daily, as of July 3rd, there have been over 80 large-scale models with parameters above 1 billion in China. In addition to the Internet giants, SenseTime, CloudWalk Technology, IFlytek and other AI track listed companies, other start-ups such as Guangnian and Baichuan Intelligence, as well as scientific research institutes such as Shanghai Artificial Intelligence Laboratory and Harbin Institute of Technology, have successively released the big models.

"We divide the big models and related products into three categories. The first category is the general big models, such as OpenAI's ChatGPT, Baidu's ERNIE Bot, and IFlytek's Spark, which belong to the general big models; the second category is the industry big models; and the third category is the application services based on the general big models or industry big models. Currently, most of the released products are concentrated in the first and second categories." Some insiders said, "From the perspective of parameter quantity, the parameter quantity of general big models such as' ERNIE Bot 'is at the level of 100 billion, and that of other enterprises or start-ups is usually at the level of 10 billion or 1 billion."

Is there duplicate construction in the "Hundred Model Battle"? Chen Duan, director of the Digital Economy Integration, Innovation and Development Center of Central University of Finance and Economics, said in an interview with Securities Daily: "In the PC era and the mobile The Internet Age, China has grown up a group of once leading Internet enterprises in the world. It is a good thing that these enterprises have actively participated in the early R&D competition of large models in response to the needs of the next round of survival competition. However, the competition in the field of general large models will produce the Matthew effect in the future, and only a few strong players can win."

Zhang Xiaorong, President of the Deep Science and Technology Research Institute, also stated in an interview with Securities Daily: There is a high threshold for training large models, and due to limitations in data and computing power, there will not be many enterprises that can truly achieve the commercialization of large models. Currently, large models in China are still in the laboratory stage, and manufacturers need more time and resources to train models, debug parameters, and so on. The market competition environment is conducive to industrial development, and ultimately, there will be waves of sand washing

Landing applications are key

From the released large models, it can be seen that there are a large number of universal large models and a few application-oriented large models. Currently, there is a gap between domestic products and ChatGPT. We need to face this gap and make key breakthroughs through scenario based traction, creating benchmarks and demonstrations to effectively and quickly narrow the gap. "Wang Peng, a researcher at the Beijing Academy of Social Sciences, said in an interview with Securities Daily.

Chen Duan told reporters, "The market demand for realizing the value of large models in specific fields and scenarios is very high, and there will be good market compatibility to accommodate more competitors. Therefore, for more participants, exploring application innovation based on large models will be a breakthrough direction

This viewpoint has also been highly recognized by track companies. Robin Lee, the founder, chairman and CEO of Baidu, said recently: "What is more important than the number of large models is the application, which is a breakthrough in the application of vertical fields. The key point of the new international competitive strategy is not how many large models there are, but how many native applications there are on the large models, and how much these applications have improved the production efficiency."

At the press conference of the listed company Jiadu Technology's "Jiadu Zhixing Transportation Model", The chairman of the company, Liu Wei, stated: The birth of the universal big model is only a starting point, and it ultimately needs to focus on specific application scenarios and solve industry specific problems. With the industry big model as the key driver, empowering industry production efficiency and service quality improvement will bring about economic and social development and profound changes in the industry. The data and feedback brought by the advantages of application scenarios will further accelerate the iteration of industry big model technology, which will also become the foundation for domestic big models Maximum competitiveness

It is understood that in fields such as government affairs, public safety, and healthcare, vertical application oriented large models are gradually being implemented. Based on its own official documents, policy documents, government affairs guidelines, and other data as professional training data, the listed company Tops has created a major model of government affairs; The MYAI big model independently developed by Myaberco focuses on the application in the vertical field, and has been applied in the construction of Big data projects in many industrial fields, such as public security, government affairs, taxation, and enterprise Digital transformation.

Disclaimer: The content of this article is sourced from the internet. The copyright of the text, images, and other materials belongs to the original author. The platform reprints the materials for the purpose of conveying more information. The content of the article is for reference and learning only, and should not be used for commercial purposes. If it infringes on your legitimate rights and interests, please contact us promptly and we will handle it as soon as possible! We respect copyright and are committed to protecting it. Thank you for sharing.(Email:[email protected])