Dozens of AI native applications are launched continuously. What signal does Baidu release?

"Without rich AI native applications built on basic models, large models are worthless." Robin Li gave a new judgment on AI trends.

Today, at Baidu World 2023, Baidu brought Wenxin Large Model 4.0, which has a comprehensive level that is not inferior to GPT4, further widening the gap with domestic competitors. "The emergence of intelligence brought by large models is the basis for us to develop native AI applications." Robin Li said.

More importantly, after Wen Xin’s words, dozens of AI native applications were released today, which revealed unusual information.

When Wen Xinyiyan App was released in March this year, he gave a prediction: the next step should be the application of large models. The previous warm-up theme of Baidu World was "Robin Li teaches everyone how to make AI native applications step by step". Combined with the series of native applications and plug-in platforms released today, Baidu is obviously sending a message to the industry: The market is starting to roll out large models Turning to volume applications, the AI ​​native era is coming.

Reconstructing products: launching dozens of applications

As the leader of China's AI, Robin Li has long said: "In the AI ​​era, every product is worth redoing." As China's first major Internet company to devote itself to AI and have been deeply involved in AI for ten years, Baidu It is becoming the first company to redo all its products.

At this Baidu World Conference, 11 new AI native applications were released. After watching the live demonstration, you will sigh insincerely: Search is no longer that search, and the map is no longer a map. Baidu seems to have It is no longer the Baidu of the past.

In the Internet era, Baidu defined search. In the past, we would not say search, but were accustomed to "Baidu". That familiar search box is a sign of the search era. But today's search with the blessing of large models, from shape to soul, is no longer the search we are familiar with.

The new search has three major characteristics: first, ultimate satisfaction, understanding you better; second, recommendation stimulation, stimulating user exploration and needs through priority recommendations; third, multiple rounds of interaction, step by step to more accurately understand user needs. You will find that the way you use search today is completely different. In Robin Li’s words: “It not only improves the quality of search, but also breaks through the application scope of search and expands product boundaries. Such AI interactive search will lead the search industry Generational changes." Baidu has redefined search for the second time. In the past, it was about searching for information, and in the future, it will be about direct services.

Baidu Map is comprehensively reconstructed from the two dimensions of "new interaction" and "new navigation". Users can have multiple rounds of conversations with the map through natural language, which not only allows the map to understand you better, but also awakens various functions and services buried in the map to help users improve travel and decision-making efficiency. In the past, maps were used for navigation. In the future, with the reconstruction of large models, more considerate services such as itinerary reminders, service reservations, and city tour guides will become a reality one by one. Baidu has redefined maps. It used to be navigation, but in the future it will be an intelligent travel guide.

The input method has also been refactored. Baidu Input Method's intelligent cloud association, cloud prediction and other functions, starting from the first word you input, there will be associated words that appear later, as well as recommended search results for you based on its understanding of you. Sometimes you only need to enter a few words With just one word, the results you want will already appear. "Super Can Write" provides various functions such as high-EQ communication, divine comments, popular titles, short video text, association continuation during input, text polishing, etc., to assist users in content creation in all scenarios and on all platforms. Obviously, Baidu Input Method is no longer an "input tool", but an "AI creation tool".

Traditional BI tools have high barriers to entry and long decision-making cycles. Baidu this time brought a new GBI - the first generative business intelligence product in China. GBI subverts traditional data analysis in three aspects: First, traditional BI can only be operated by professionals, while GBI can directly understand the problem and execute it in real time; second, GBI provides a convenient access method, and enterprises can access data. Questioning and analyzing any data eliminates the need for manual cross-database and cross-table analysis; thirdly, GBI also has learning capabilities, and companies can inject professional knowledge in the industry to make it an industry expert. It can be said that Baidu has redefined BI in the AI ​​era, allowing enterprises to stay one step ahead in the fierce competition in shopping malls such as battlefields.

Obviously, Robin Li is turning his prediction of trends into the company's actions, and he is moving very quickly, truly realizing the reconstruction of Baidu products such as Baidu's new search, new library, new network disk, and new map, and creatively Developed Baidu GBI, a new AI native application, becoming the first technology company to use AI to reconstruct products.

In Dongdong’s view, the release of dozens of native applications is the reconstruction of dozens of products and the redefinition of dozens of product categories. This is like the subversion from horse-drawn carriages to cars, from feature phones to smartphones. It is an intergenerational revolution, not a simple upgrade.

Reconstructing value: rewriting the direction of AI competition

Beginning in the first half of this year, the entire industry began to roll out large models. In fact, for users, the ultimate goal of large-scale AI model development is to implement and transform it into a tool that can be used by ordinary people.

In the past few months, every product at Baidu has been restructured in full swing, and the understanding of native AI applications has been deepened, in order to implement large models and realize user value.

Take Baidu Wenku as an example. This is the most thoroughly reconstructed product. While the product is reconstructed, it also realizes value reconstruction.

In March this year, after Wenxinyiyan went online, the Wenku team started taking action. "We are not combining with AI, we are reconstructing." Wang Ying, vice president of Baidu and head of interactive entertainment and vertical categories, said that the reconstruction process is not based on how to upgrade Wenku's original products, but from the user's perspective. Let’s look at the pain points in the whole process of document writing. We will provide solutions to the pain points one by one and reconstruct the process of user document creation.

After months of painstaking reconstruction, in August Baidu Wenku announced that it would be restructured based on Wenxin Yiyan and upgraded to a "one-stop intelligent document platform", opening up "PPT intelligent generation, document intelligent generation, intelligent editing, and intelligent assistance" Read "The Four Major AI Capabilities.

Traditional document writing is difficult to write, edit, have multiple platforms, and organize. The new Baidu Wenku solves these pain points. Through one-stop intelligent document full-process services, users do not need to create documents, edit documents, and convert documents. Document formats can be switched between multiple platforms, and the entire process from searching for information, finding ideas, writing documents, modifying and editing can be completed in one stop.

The reconstruction of Baidu Library can be said to be a model of reconstruction in the AI ​​era.

The first is the reconstruction of the product level. "Based on the four major capabilities of the Wenxin model, it is no exaggeration to say that Baidu Wenku has fundamentally revolutionized its own life and completed the evolution from a content tool to a productivity tool." Robin Li emphasized.

Secondly, it is the reconstruction of user value level. Baidu Wenku will completely subvert users’ office and creative habits. This is a bit like in the era of feature phones, where mobile phones were used as communication tools for making calls and sending text messages. In the era of smart phones, mobile phones are the terminal that connects people to smart life. People's lifestyles are gradually changing, and mobile phones are inseparable from food, clothing, housing and transportation. "It has become a new product, and users have new expectations for it." Wang Ying said.

Whether a product can be reborn in the AI ​​era depends on whether it can create new value for users. Baidu Wenku has been online for more than a month without large-scale publicity, and has accumulated more than 10 million users and nearly 100 million uses. It has generated more than 2 million PPTs and generated more than 20 million content. The key point is that the new features are widely welcomed and praised by users. "We hope to be truly valuable, solve users' pain points, and make users like us. Our value can be seen from the data." Friends around her constantly reported that Baidu Wenku was used to solve practical problems in work and life, which made Wang Ying very excited.

Reconstruction is a wonderful chemical reaction, and some unexpected changes are also happening with it. Wenku was not considered a core business within the Baidu system before, but product reconstruction has brought about value reconstruction, which has increased its importance to users and has a closer relationship with users. As Baidu fully enters the AI ​​era, the importance of Baidu Library in Baidu's business system has also been enhanced.

Baidu Wenku is one of many reconstructed products. The success of Wenku heralds the rewriting of the direction of AI competition. Everyone knows that AI brings huge industry opportunities, but where are the opportunities? It is the realization of user value. What realizes user value is one AI native application after another. Which native application can achieve greater user value will stand out in the next era.

This time Baidu World released the reconstructed application in "batch", leading the new direction of the AI ​​competition: from rolling large models to the stage of rolling applications. As Robin Li said: "We will all enter an era of native AI, an era of human-computer interaction through prompts."

Reconstructing the future:

Native applications emerge, fast fish eat slow fish

Everyone still remembers that AlphaGo defeated the Go champion in 2016 and attracted the attention of the world, right? However, the heat of this matter soon passed and disappeared. AlphaGo is a small bright spot in the history of AI evolution, but it is not a key node that affects the industrial process. On the contrary, once Wenxinyiyan APP, Chat GPT and midjourney were launched, they became popular all over the world due to their strong practicality, driving the entire industry forward.

History and reality both prove a logic. The ultimate goal of all technologies, including AI, is to serve a better life for mankind and the development of society. AI that cannot be implemented is just a castle in the air. AI is more than just playing chess and chatting with humans. It is more than just entertainment. More importantly, it can be implemented in production and life, making life more convenient and upgrading industries to improve efficiency.

As early as March this year, Robin Li said that "the real opportunity brought by large models is application." In the past six months, Robin Li has emphasized the importance of application in large models on multiple occasions: he has defined three conditions for native AI applications: natural language interaction, full use of the four capabilities of understanding and generating logical memory, and interaction not exceeding Two level menu.

In the mature stage of the industry, bigness is the goal pursued by enterprises, and the competitive logic at that time is that the big fish eats the small fish. In the transformation period of the industry and the arrival of a new era, speed is the first factor. The fast fish eats the slow fish.

Baidu is the first major Internet company to release large models, and Wen Xinyiyan is among the first batch of large models to obtain licenses and iterates rapidly. This time, Wenxin Large Model 4.0 is comparable to Chat GPT, leading all domestic large models. When it comes to rolling out large models, Baidu is the fastest with its decade of experience in AI.

Thanks to Robin Li's forward-looking judgment, the entire Baidu has been rapidly changing from the inside out in the past few months, defining the form of "AI native applications", leading Baidu to rapidly explore, innovate and reconstruct. This time Baidu has led the industry in a new direction with "fast" - native applications.

Of course, Baidu not only wants to be fast, but also hopes to drive the industry, partners, and customers to be fast.

"Large models will open up a prosperous AI native application ecosystem," Robin Li believes that plug-ins are a special kind of AI native applications with the lowest threshold and the easiest to get started, allowing developers and entrepreneurs to quickly join. in ecology.

As a new form of the AI ​​era, plug-ins are the starting point for enterprises to catch up with the train of the AI ​​era and seize the opportunity. A month ago, Baidu launched the Lingjing plug-in platform. Currently, 27,000 developers have applied to join it, covering many fields such as law, workplace, and learning.

In order to lower the threshold for AI native application development, Baidu Smart Cloud has released the "Qianfan AI Native Application Development Workbench", which integrates common patterns, tools, and processes for developing large-model applications into a workbench to help developers focus on own business without involving unnecessary energy in the development process. In addition, Baidu Smart Cloud also launched Qianfan AI Native App Store, China's first one-stop AI native application store for enterprise customers, to accelerate the commercialization of AI native applications.

If the large AI model is the wheel of the times, then the AI ​​native application ecosystem is the engine of the times. From Baidu's own restructured products to the plug-ins that developers are passionate about, this Baidu World indicates that engines are being ignited and AI will become an important driving force for the next round of economic growth.

【Conclusion】

AI is reconstructing products, reconstructing value, and also reconstructing the future.

"The new world and the new future will be generated through prompts from every enterprise, every developer, and every user." Robin Li issued an invitation at this Baidu World: "The future will be generated by you!" The future is created by us together!”

The above is the detailed content of Dozens of AI native applications are launched continuously. What signal does Baidu release?. For more information, please follow other related articles on the PHP Chinese website!