There is big news in the field of AI search engines. On December 25, a news that "Douyin tests the 'AI search' function, and can summarize the answer from the intelligent search of the Douyin app" sparked heated discussions on the Internet. According to a number of ** reports, Douyin recently tested a new feature in the app, called "AI search". This is an AI-based search Q&A service that provides users with answers.
In this regard, the relevant person in charge of Douyin told a reporter from China ** Daily that the relevant functions are still being tested and are not yet mature. The reporter tried to use this feature, but found that it was not displayed in the Douyin app. Industry insiders said that the feature may be in the grayscale test and has not been fully launched.
Beta: "AI search" function.
It is understood that the previous "AI search" function was located in the search interface of the Douyin app, alongside other search services such as goods, **users, **, etc. Users only need to enter relevant questions or information in "AI Search", and they can get the answers generated by "AI Search" for users through intelligent search technology.
According to public information, since the beginning of this year, major domestic apps have successively launched AI search Q&A services.
* On the original search function, the e-commerce search guide method has been iterated and the "**Ask" has been launched, which has been experienced by more than 5 million people since the launch of the internal test in September this year. , Xiaohongshu and other apps have also launched AI intelligent search. Among them, Xiaohongshu tested the AI chatbot "D**Inic", which can be combined with Xiaohongshu's own product positioning to focus on chat recommendation services in tourism, daily life, food and other aspects.
The market has broad prospects.
From the perspective of the industrial chain, the latest research report of the Toubao Research Institute shows that the industrial chain of China's AI search engine industry can be divided into three links: upstream (algorithm research and development and basic software and hardware), midstream (development and operation of AI search engine platform) and downstream (application scenarios).
Specifically, the upstream links involve machine learning algorithms, artificial intelligence frameworks, computing equipment and big data collection, etc., and Tencent and Alibaba have strong competitiveness in this link. The midstream link covers search technology, natural language processing technology, knowledge graph, recommendation system, etc. The enterprises in this link are mainly search engine operators, including, Sogou, 360 search, etc. The downstream involves a wide range of application scenarios, including but not limited to finance, healthcare, education, e-commerce, advertising, etc. Downstream enterprises usually apply AI search engine technology to actual business scenarios to improve user experience and efficiency.
As for the market size, the above research report mentioned that the size of China's AI search engine market will reach 79 in 20223.5 billion yuan, with the gradual promotion of AI search engine products and users' ardent expectations for them in various application scenarios, the market development speed will continue to rise in the next 5 years, and is expected to reach 329 by 20273.5 billion yuan. This means that the market size will grow at a CAGR of around 3293%。From the perspective of the competitive landscape, the competition in the AI search engine market will become increasingly fierce in the future, and giants such as Tencent, and Alibaba are expected to form a competitive pattern.
According to the research report of Guohai**, at present, the wave of AI search engine upgrades has arrived, the technology giants have a comprehensive layout, and the middle and downstream manufacturers are blooming. Microsoft, Google, and Meta are the main ones abroad, and Tencent, Alibaba, Huawei, etc. These companies not only have sufficient computing power support, but also have excellent talent teams, years of algorithm and data accumulation, and have natural advantages in the development and application of large models.
In addition to cloud vendors, there are also optical communication vendors, data service providers, and computing power-related equipment vendors, which will greatly benefit from the more computing resources and data requirements brought about by the development of large models. In the midstream, there are companies such as SenseTime, iFLYTEK, Megvii, and Tors, which have been deployed in subdivided fields for many years, and also have certain computing power reserves, and are rich in related technologies and data reserves. It is recommended to pay attention to industrial companies with office software, e-commerce, media, marketing, education, games, virtual humans, etc.
Editor-in-charge: Wang Shidan |Review: Li Zhen |Supervisor: Wan Junwei.
*: China ** newspaper).