July 18, 2023, the current artificial intelligence market is in full swing driven by large models and generative AI, but looking back at the market situation in 2022, the landing has slowed down and negative growth has become the core keywords. IDC recently released the "Chinese Intelligent Software Market Share 2022" report, the market size of artificial intelligence software in 2022 will be 30.73 billion yuan, down 6.9% year-on-year compared with 2021.
Computer vision AI market
In 2022, the market size of the computer vision market reached 12.30 billion yuan, a year-on-year decrease of 19.1% compared with 2021. Among them, SenseTime has been at the top of the list for the past 5 years; It is followed by Hikvision, Innovation Qizhi, Megvii Technology, Yuncong Technology, and Smart Eye Technology. In addition, although cloud service vendors are not represented in the pie chart, their market size has become more and more considerable, becoming an important force in the visual market. In addition to the top vendors, vendors that provide cloud-edge-end solutions and vendors that focus on industry application scenarios such as quality inspection and inspection are also important components of the market.
Speech Semantic AI Market
In the speech semantic AI market, iFLYTEK is still at the top of the list, but Baidu Intelligent Cloud and Alibaba Cloud are constantly narrowing the distance between them and iFLYTEK. In addition to the top manufacturers tracked this time, there are many vendors that cultivate in this market segment, such as HUAWEI CLOUD, Tencent Cloud, JD Technology, and Chaoyi Technology's intelligent customer service; Puqiang Information, Soundcom Technology, Yunzhisheng, Intelligent Voice and so on. In this market, medium and long-tail vendors occupy 59% of the market share with decentralized application scenarios or solutions.
Machine learning platform market
In 2022, the market size of machine learning development platforms reached 3.54 billion yuan, down 4.9% year-on-year compared to 2021. The market shows a significant scale effect of the top manufacturers, with the top 5 manufacturers accounting for 69.9% of the market share. Due to the technical threshold of the machine learning development platform and the many challenges of landing, the market has not shown a state of blooming. Of course, with many startups entering the big model market, the market structure of machine learning platforms may be redistributed in the future.
If the overall AI market participants are divided into several groups, we can see that established listed companies such as Hikvision and iFLYTEK can still achieve small growth in the declining market; Growth start-ups such as Innovation Qizhi and Fourth Paradigm can achieve rapid growth; Large cloud vendors have seen significant growth; Most other manufacturers showed negative growth. From the perspective of market performance, enterprises with deep customer accumulation in specific markets for many years are more able to resist external adverse factors (such as sudden emergence of the epidemic, economic downturn, etc.). Growing companies, on the other hand, can either continue to deliver clearly leading technologies or find tracks with steady growth potential, otherwise it will be difficult to continue the growth rate of the past few years.
In the past year, the lack of a strong driver in the artificial intelligence market has led to the overall market declining for the first time in nearly a decade. Fortunately, the boom of ChatGPT and large models driven by OpenAI has pushed AI back to the focus, although it may not be able to achieve large-scale commercial revenue within this year, but its breadth of popularity, speed of promotion, and depth of penetration are expected to drive the AI industry to quickly cross this downward cycle and return to the era of high growth.
Lu Yanxia, research director of IDC China, said that the technological breakthrough and landing speed of large models and generative AI is beyond imagination, and this wave of AI ubiquitous will cross the technology bubble period at a faster speed than the previous wave and enter the landing stage, and will also quickly eliminate manufacturers that do not have real strength. Whether self-developed or not, IT industry players need to ensure that their products/solutions can provide this part of the capability.
(Liu Liqing)