Skip to content
After the generational breakthrough of large language models, several search engines both domestically and internationally are trying to integrate them, and the changes brought to user experience are evident.
For example, ChatGPT has already been integrated into search engines. Baidu Search also began internal testing of “AI Partner” in early May, where the AI automatically generates answers after users input questions, along with citations of information sources. Especially in scenarios such as consumer decision-making, where users need to gather information on multiple products for comparison, AI dialogue can consolidate different product information based on comparison dimensions, making it easier for users to choose and saving a lot of time.
This is a powerful rebuttal to those who still doubt that search engines will become obsolete—after algorithm recommendations became popular, such doubts arose frequently.
As an application driven by AI technology since its inception, search engines are being given new vitality by artificial intelligence. Especially as large language models gradually mature, a noticeable change is the significant improvement in the reasoning ability of search engines, achieving generational breakthroughs in reasoning, organization, and even creative capabilities, along with leaps in service capacity and efficiency.
The development process of search engines is increasingly understanding users, providing richer information and content. From the early platform continuously optimizing the ranking order of search results to later directly structuring and presenting the information users want, every change in user experience brought by search engines also enhances user scale and commercial value.
Today, search engines are once again at the brink of transformation. As various companies compete for the first-mover advantage brought by large model technology, what changes will occur in the platform’s service capabilities and commercialization as AI becomes accessible to everyone? What new variables will emerge in industry competition? At the 2023 Wanxiang Conference, Baidu provided its own answers.
Large Models Change the Future Content Ecology
New technologies with transformative capabilities often change our lives and production in ways and speeds we least expect.
Today, artificial intelligence technology possesses revolutionary capabilities, changing everything at an accelerated pace. Big data, big computing power, and large models allow machines to learn comprehensively, giving AI the ability to learn and generate independently. This not only represents a significant change for search engines but also fundamentally alters the existing content creation and distribution ecology.
On one hand, there is a change in human-computer interaction methods. In the past, users searched for information and content in the Baidu search box or explored the unknown in the information flow of the Baidu APP. In the future, with the dual capabilities of “semantic search technology + large language models,” users can obtain information at a lower cost.
At the 2023 Wanxiang Conference, He Junjie, Senior Vice President of Baidu Group and General Manager of Baidu Mobile Ecosystem Group (MEG), revealed that Baidu Search’s “AI Partner” and “AI BOT” built with large models are undergoing comprehensive internal testing. Specifically, the “AI Partner” accompanies users like a friend, engaging in voice conversations, reducing the input expression threshold and cost for users. Meanwhile, creators, businesses, institutions, services, and brands can establish their own “AI BOT” based on existing knowledge, content, services, and experiences, turning it into a virtual expert that is available 24/7. At the same time, the Baidu APP will leverage large language models to provide users with personalized homepages, allowing users to meet different needs through five major scenarios: “look, search, ask, buy, listen,” truly achieving “one person, one world.”
On the other hand, there is a transformation in content production methods. From the explosive popularity of AI painting applications like Wenxin Yige, Midjourney, NovelAI, to the successive launches of large language models like ChatGPT and Wenxin Yiyan, the content production field has entered an era of “big explosion” in application tools. The number of AIGC applications is continuously increasing, producing content at hundreds or thousands of times the speed and at costs of one-hundredth or even one-thousandth that of human production, bringing about a revolution in content production.
Currently, Baidu has 450,000 AIGC creators who have produced over 7 million pieces of content, with a cumulative distribution volume exceeding 20 billion. To better empower creators, Baidu announced an upgrade of its AIGC creation tools at this year’s Wanxiang Conference and launched the “AI Co-Creation Plan,” providing creators with more comprehensive AI creation solutions such as AI notes, AI BOTs, and AI digital humans.
Baidu is actively participating in the wave of content ecology transformation brought by large models. From the information released by Baidu at the 2023 Wanxiang Conference, its ambition in this transformative era is to reconstruct the content ecology using AI-native thinking and concepts, improving the interaction methods and content creation methods of every product in the Baidu mobile ecosystem, seizing the initiative in the transformation.
Baidu’s Action to Reshape Mobile Ecology Has Already Begun
According to Baidu’s plan, in the near future, well-known products such as Baidu Search, Baidu APP, Baijiahao, and Baidu Wenku will all feature new human-computer interaction methods and AI creation forms.
This is a well-prepared battle. In this wave of artificial intelligence, Baidu’s mobile ecosystem can act quickly due to its continuous investment and innovation in artificial intelligence for over ten years. In fact, through ongoing investment in artificial intelligence, Baidu’s actions to reshape the mobile ecosystem have already begun.
Although the Baidu homepage has existed as a simple search box for over twenty years, the technology behind the search box has continuously innovated with changing user demands. Baidu Search has made significant investments and innovations in products and technology, especially in AI technology, and the changes in user experience are evident.
As one of the earliest technology internet companies in China to engage in artificial intelligence, Baidu has invested over 100 billion yuan in research and development, having long predicted the trend of large models and AIGC. Four years ago, Baidu launched the Wenxin 1.0 large language model with over 100 million parameters and continuously refined Baidu’s large model technology through key applications in search; two years ago, Baidu realized that the development direction of artificial intelligence was shifting from recognition to generation and began to allocate resources towards AIGC.
With the support of artificial intelligence technology, a noticeable change is that Baidu Search can now understand not only text information but also various forms of information such as voice and photos, allowing users to input questions in a more conversational manner. With the aid of AI technology, the search engine has been equipped with senses and a brain, capable of “understanding,” “seeing,” and “hearing” these demands.
The enhancement of user experience has also driven continuous growth in Baidu’s mobile ecosystem users. Baidu’s first-quarter 2023 financial report shows that in March, the monthly active users of the Baidu App reached 657 million, a year-on-year increase of 4%. The number of search queries and information flow distribution on Baidu’s mobile end steadily increased. Data from monitoring agency SimilarWeb indicates that Baidu is not only the number one search engine in the Chinese market but also ranks first in overall webpage visits in China.
It is foreseeable that as AI-native thinking and concepts are applied more broadly in products like Baidu Search within the mobile ecosystem, more usage scenarios will be innovated, providing users with a better experience, which will inevitably drive the user scale and traffic of Baidu’s mobile ecosystem to continue growing.
Greater Imagination Lies in Business Models
New traffic entry points, interaction methods, and content creation methods will inevitably lead to the reshaping of business models. This also means that with artificial intelligence technologies represented by large models, search engines and content distribution will possess greater imaginative space in the commercial field.
On one hand, new technologies will drive the marketing industry to find new solutions, achieving cost reduction and efficiency enhancement. For example, Baidu’s digital human Duxiaoxiao collaborated with OnePlus to generate the country’s first digital human 3C evaluation video through AIGC capabilities. This means that creators can utilize AIGC to fulfill commercial orders and continue to monetize, while enterprises can use AIGC to replace human customer service, resulting in a 100% increase in sales leads in certain scenarios.
On the other hand, AI will drive the content ecology to grow new business cycles. Baidu’s intelligent e-commerce has emerged as a new species in the new content ecology. Unlike shelf e-commerce and content e-commerce, Baidu’s intelligent e-commerce relies on AI to recombine users, media, products, and services, launching a new model of “search-push-browse integration,” which has gained recognition from users and businesses. The latest data shows that Baidu e-commerce’s total GMV has increased by 152% year-on-year, with monthly trading user numbers growing by 100%, a 25% increase in repurchase rates, and a 53% increase in the number of businesses joining Baidu e-commerce.
At the same time, leveraging AIGC technology, Baidu e-commerce has launched a digital human live streaming platform, allowing businesses to create videos in just five minutes and use AI to assist in script creation, significantly reducing content production costs.
With the support of large models and AIGC technology, whether for advertisers or e-commerce merchants, the future will belong to those who possess the best methods of communication with customers. Institutions predict that in 2022, 2% of marketing information from large organizations was generated by AI, and by 2025, this figure will rise to 30%.
From this perspective, the reconstruction of business models in the content ecology by artificial intelligence is just beginning, with even greater imaginative space ahead.
While many unknowns still exist in the future, the trend of reconstruction has already been very apparent in the past decade of the mobile internet wave: when the iPhone emerged as a new species, it not only brought updates to hardware devices but also established a completely new mobile internet ecology, giving rise to various applications and business models.
It is foreseeable that as artificial intelligence descends from the pedestal of technology and deeply integrates into practical applications, our modes of thinking, social life, and business models will undergo a more thorough transformation. Specifically, in the Baidu mobile ecosystem, as large models and AIGC are applied more widely, the business and commercial models of Baidu’s mobile ecosystem will also be completely renewed.
Special Planning
