Baidu Achieves New Growth Through AI-Native Applications
-
On November 21, Baidu released its Q3 financial report, showing that its revenue reached 34.447 billion yuan, with net profit attributable to Baidu (non-GAAP) hitting 7.3 billion yuan, a year-on-year increase of 23%. Among these figures, Baidu Core revenue for the quarter was 26.6 billion yuan, with net profit attributable to Baidu Core (non-GAAP) reaching 7 billion yuan, up 21% year-on-year. Baidu Core's online marketing revenue grew to 19.7 billion yuan.
Baidu's Q3 revenue and profits both exceeded market expectations, aligning with the optimistic forecasts from various sectors. International and domestic investment banks widely believe that the sustained significant growth in Baidu's core businesses like advertising and cloud services is unsurprising, given its continuous development of AI-native applications.
Bank of Communications International stated, "We believe that ERNIE Bot, leveraging Baidu's traffic, search product advantages, and technical/financial support, is poised for rapid growth. We are optimistic about its potential to restructure Baidu's product portfolio and drive revenue."
CITIC Securities was even more straightforward, "After the Baidu World Conference, AI's transformation of the company's advertising and cloud computing businesses has deepened, potentially increasing user engagement, ad pricing, cloud infrastructure usage, and cloud-native business growth. We remain positive about the company's fundamental recovery and the investment opportunities brought by new AI businesses, maintaining a 'Buy' rating."
Even Morgan Stanley couldn't help but praise Baidu, calling it "China's best AI internet company and a pioneer in driving AI breakthroughs in China."
The widespread optimism about Baidu's performance stems from one key factor: while others are still developing AI large models, Baidu has already taken the lead in launching a series of AI-native applications.
Baidu AI native application's user base and API calls have shown significant growth.
First, let's talk about the growth in user base.
Data shows that compared to 2022, Baidu's AI native application user base grew by 50% in 2023. Among them, the new user growth rate reached 60%, while the active user growth rate was 45%.
Taking the Wenxin Big Model 4.0 as an example, Wenxin Yiyan officially started its invitation test on August 16, 2023, and was the first to fully open to the public on August 31. Within just 12 hours of its release, the Wenxin Yiyan APP topped the Apple Store's free app rankings. Within 24 hours of its release, Wenxin Yiyan answered over 33.42 million questions from users.
On November 8, Baidu's knowledge-enhanced large language model key technology won the "World Internet Conference Leading Technology Award." As of now, Wenxin Yiyan has reached 70 million users, 4,300 scenarios, and 2,452 applications.
In addition, since the launch of Baidu Wenku's new AI features, the cumulative number of users has exceeded 13 million, and the cumulative number of feature uses has surpassed 100 million. Even the "Super Writer," an all-scenario AI creation assistant launched via Baidu Input Method, has performed exceptionally well: by providing features such as high-EQ communication, divine comments, and associative continuation, it has accumulated over 100 million requests, with a user like rate exceeding 80%.
Next, let's discuss the growth in API calls.
In addition to the growth in user base, the API calls for Baidu's AI native applications have also seen a significant increase in 2023. Data shows that API calls grew by 60%, reaching millions of daily calls.
Among these, the call volume for voice recognition and natural language generation functions has grown most significantly, with a growth rate as high as 80%. Specifically, the call volume data, as Robin Li mentioned, "might be larger than the combined call volume of over 200 domestic companies," and continues to show an exponential growth trend.
This series of significant data growth clearly indicates that Baidu's AI-native applications have gained widespread favor in the market.
It also echoes a statement made by Robin Li after the release of ERNIE 3.5: "ERNIE 3.5 (Wenxin Large Model 3.5) has received widespread recognition from our cloud customers, AI developers, and industry experts. We are reconstructing Baidu's products and services with an AI-native mindset to create innovative experiences. At the same time, we are using large models to support various enterprises in seizing this opportunity.
Overall, Baidu is committed to building a new growth engine based on generative AI and large language models to drive sustainable long-term growth."
The popularity of AI-native applications has had a profound impact on Baidu's core businesses, advertising, and cloud services.
On October 17, "Baidu World 2023" was held at Shougang Park in Beijing. At the time, Robin Li delivered a speech titled "Step-by-Step Guide to Building AI-Native Applications," released the Wenxin Large Model 4.0, and introduced over ten AI-native applications, including new cloud storage and new maps.
In terms of advertising business
This wave of major actions before the Q3 earnings release has brought additional revenue to Baidu's advertising and cloud businesses, with an effect far exceeding the simple sum of one plus one.
As is well known, large models possess powerful data processing and pattern recognition capabilities, enabling effective analysis of massive data, thereby creating new growth opportunities for advertising and cloud businesses.
In the advertising sector, large models help advertisers more accurately identify potential customers and improve ad placement effectiveness. Through in-depth analysis of user behavior and interests, large models can generate more personalized ad recommendations, increasing click-through and conversion rates. This not only provides advertisers with more efficient returns on investment but also enriches users' consumption experience.
In the cloud business sector, large models offer more robust computing and storage capabilities. By processing and analyzing large-scale data, large models enable businesses to better understand market trends and optimize operational processes.
At the same time, large models provide developers with more efficient development tools, reducing development costs and further driving the growth of cloud computing businesses.
Taking Baidu Intelligent Cloud Qianfan as an example, Qianfan currently serves over 20,000 clients, with nearly 10,000 monthly active enterprises on the platform, covering nearly 500 scenarios.
Leading clients across various industries, including Sany Heavy Industry, Guoneng Yulin Energy, Hebei Expressway Group, Postal Savings Bank of China, Kingsoft Office, Kingdee, and iSoftStone, are exploring the path of industrial innovation with large models through Baidu Intelligent Cloud Qianfan.
During the World Conference, Baidu announced a comprehensive upgrade of its "Cloud Intelligence Integration" strategy, offering full-stack service solutions to meet clients' needs for large model implementation.
For AI-native application development, Baidu launched the "Qianfan AI-Native Application Development Workbench" to accelerate the deployment of enterprise AI-native applications.
Additionally, Baidu released an AI-native application store and a comprehensive large-model ecosystem support system.
Empowering partners' business growth, co-building and sharing a prosperous large model industry ecosystem.
Perhaps it is precisely these continuous "surprises" that have enabled Baidu, as one of China's earliest internet giants to invest in AI, to consistently take the lead in the AI-native application competition phase. It is both a pioneer and a frontrunner, continuously achieving growth against the odds.
Valuing technological accumulation and ecosystem expansion, Baidu's steady progress is a natural outcome.
At its core, Baidu's leading position in the AI technology field is primarily reflected in its technological accumulation and expansion.
1. Number of Patent Applications: According to public data, Baidu has filed over 22,000 patent applications in the AI field, with more than 4,600 authorized patents. These patents cover various areas such as speech recognition, natural language processing, computer vision, and deep learning, showcasing Baidu's comprehensive and in-depth R&D capabilities in AI.
2. Breakthroughs in Core Technologies: Baidu has achieved remarkable results in AI technology R&D. For example, its self-developed "PaddlePaddle" deep learning platform has become China's first independently developed deep learning framework with proprietary intellectual property rights, providing robust technical support for global developers. Additionally, Baidu has made a series of groundbreaking achievements in speech recognition, natural language processing, and computer vision, such as achieving a 98% speech recognition rate and leading globally in natural language processing technology.
3. Building an Innovation Ecosystem: Baidu actively promotes the construction of an innovation ecosystem by collaborating with various enterprises, research institutions, and universities to create an open, shared, and innovative AI ecosystem. This ecosystem not only accelerates the rapid development of AI technology but also provides strong support for China's technological innovation.
4. Expansion of Application Scenarios: Baidu's intelligent AI has made significant progress in areas such as speech recognition, image recognition, machine translation, natural language processing, and intelligent recommendation systems, enabling AI technology to better serve people's lives and work. For instance, through deep learning technology, Baidu's intelligent AI has achieved a speech recognition accuracy far surpassing traditional methods, accurately converting human speech into text. This has made voice interaction a reality, greatly enhancing convenience in daily life and work.
In summary, with the continuous advancement of artificial intelligence technology and the expansion of application fields in the future, Baidu's well-prepared AI-native applications will undoubtedly bring more new sources of incremental growth to the company.
At the same time, as Baidu continues to increase its investment and innovation in AI-native applications, it will inevitably drive the continuous development of artificial intelligence technology in China and even the world. The intelligent future may not be far away.