Absorb 20 billion yuan in gold annually! The annual salary of relevant talents exceeds 600000 yuan! Another new trend?
Recently, the first batch of large model products registered through the Interim Measures for the Management of Generative Artificial Intelligence Services in China have been launched and officially opened to the public. Large model technology refers to the use of large-scale neural network models and language based interaction to complete various artificial intelligence tasks. Currently, large model technology is rapidly iterating globally and has applications such as ChatGPT. So, what is the current situation of the development of the domestic large model market?
"Hundred Model Battle" Racing Horse enclosure
Centralized launch of large-scale artificial intelligence models in China
A startup enterprise in Beijing that independently develops large-scale models has launched a large-scale model application that has been open to the public for nearly 10 days. The staff is still optimizing and improving the content generated by the large-scale model. Currently, a large number of Q&A sessions are focused on text creation, education, and other fields.
Zhang Peng, CEO of a certain artificial intelligence big model enterprise: It is itself a big language model, so its language ability is stronger. We estimate that it is equivalent to the level of a bachelor's degree graduate, and the ability in mathematics and science is close to the level of middle school. We will upgrade the model in a few months.
According to statistics, 11 large models in China have successively passed the Provisional Measures for the Management of Generative Artificial Intelligence Services, with 5 in Beijing and 3 in Shanghai being the first to go online, and 2 in Guangdong and 1 in Anhui also gradually opening up.
The reporter learned that the latest general model launched in China is more like an infrastructure, which can incubate vertical models in more professional fields and serve the digital transformation of various industries.
Industry insiders say that the current investment cost of general large models is very high and needs to be effectively combined with specific application scenarios. The related industries are still in the early stages of development.
Nie Junfeng, a researcher at the China Banking Research Center of Central University of Finance and Economics: We are currently in a stage of market competition, industry rules have not yet been established, and market patterns have not yet fully formed. People call it the "Hundred Model Wars", and since last year, there have been nearly 100 financing cases in the large model industry in China. In terms of investment valuation for the entire industry, the new investment in this industry in 2023 is estimated to be between 15 billion to 20 billion yuan.
Scarcity of large model talents and rising salaries
The market for computing power algorithms is active
The heating up of the large model market is also driving the upstream and downstream industrial chains, including data, computing power, and algorithms. Talents in related fields, such as large model engineers, are also in short supply in China, and their income levels continue to rise.
Liu Xinyang is a large model engineer whose main job every day is to train large models for various industries. Prior to this, he worked in search engine development. In 2022, he transitioned to research and development of large model technology.
Liu Xinyang, a big model engineer for a certain artificial intelligence enterprise: The big models I trained include the financial industry, real estate industry, medical industry, and water conservancy. After all, large models are a newly emerging technology that requires some training before they can be used.
According to statistics from relevant institutions, the average annual salary of domestic artificial intelligence engineers and algorithm scientists has exceeded 600000 yuan, and there is a strong demand for talent in large model technology research and development by enterprises, with significant gaps in some core areas.
Liu Huifu, Chief Technology Officer of a certain information technology group: Experts who truly understand the industry are very needed, and those who understand artificial intelligence, algorithms, and models are also very needed, because the result of each round of artificial intelligence development is to enable those who are good at using tools to survive better.
The heating up of the big model market is also driving the computing algorithm industry, including chips, computing clusters, servers, communication network hardware, as well as software such as cloud computing, databases, and virtualization functions for big models. Computing infrastructure and services are essential for training large models, as their performance, quality, and stability determine the effectiveness of model training.
In the international market, the price of a chip specifically designed for training large artificial intelligence models has even been hyped up to $45000, equivalent to over 300000 yuan in Chinese yuan. The independent research and development of chips in related fields in China is also accelerating.
Liang Zheng, Vice Dean of the Institute of Artificial Intelligence International Governance at Tsinghua University: Computing power is a relatively high threshold for the implementation of large models. At this level, the country is already coordinating and deploying, which is the construction of our national computing power network. There are some computing centers in various places, which used to be mostly CPUs. Now, we are adding intelligent computing GPUs.