Shulou(Shulou.com)11/24 Report--

"the emergence of intelligence brought about by large models is the basis for us to develop AI native applications." On October 17, Robin Li said on Baidu World 2023. On the same day, Robin Li delivered a speech with the theme of "hand-in-hand teaching you to do AI native applications", released Wen Xin Da Model version 4.0, and brought more than a dozen native AI applications, such as new search and new maps.

At the conference, Robin Li announced the official release of Wen Xin Da Model 4.0 and opened the invitation test. He said that this is by far the most powerful literary heart model, realizing a comprehensive upgrade of the basic model in understanding, generation and logic.

Both the collection and memory ability have been significantly improved, and the comprehensive ability is "not inferior to that of GPT-4". According to Li Yanhong, Wen Xin 4.0 also began to invite testing at the same time, and the audience scanned the QR code of Wenxin card, logged on to the official website of Wenxin Yiyan or downloaded the latest version of Wenxin Yiyan APP, you can experience the professional version of Wenxin Yiyan; in addition, enterprise customers can also apply to test Wenxin 4.0 API through Baidu Intelligent Cloud Qianfan model platform.

He showed more than a dozen AI native applications based on Wen Xin word reconstruction, such as Baidu search, Ruiliu, map, web disk, library, and so on, hoping to expand everyone's imagination and "inspire everyone to make more amazing AI native applications".

The comprehensive release ability of the strongest literary heart model 4.0 is not inferior to that of GPT-4.

In Li Yanhong's view, the birth of AI native applications benefits from the four core competencies of understanding, generation, logic and memory of big models. Baidu's AI native applications are also developed based on Wen Xin. "these capabilities are not available in the past, so they can open up unlimited space for innovation."

Based on Wen Xin Da Model 4.0, Robin Li demonstrated the characteristics and application scenarios of the four capabilities in turn. In terms of understanding ability, by asking about the case of provident fund loan policy, he demonstrated the comprehension of Wen Xin's words on complex prompts such as disorder, vague intention, subtext, and so on. For example, "working in Beijing" is equivalent to "paying provident fund in Beijing" and so on. "Today, every word you say is likely to be understood."

In terms of generating ability, Li Yanhong shows how Wen Xin's words can quickly generate a set of advertising posters, five advertising copywriters and a marketing video based on a material picture in just a few minutes. According to reports, based on this series of capabilities, Baidu has launched the AIGC marketing creative platform rudder, so that "a person becomes an AI marketing team."

At the same time, he also demonstrated the logical ability of the big model by solving math problems and summing up knowledge points, and reflected the memory ability of the big model through thousands of words of novel writing, characters and plot settings. and digital doctors help patients interpret drug instructions to show the comprehensive application of the four abilities.

"the previous demonstration reflects the progress of the literary heart model in the four abilities of understanding, generation, logic and memory, which are the basis for the survival of all AI native applications." Li Yanhong said.

More than a dozen native AI applications have been released

The rich native application of AI is the value of the big model. At the conference, Robin Li announced that "our search, stream, map, web disk, library, etc., will all be presented to you with a new look," and said that the purpose of sharing these applications is to expand their imagination and inspire more people to make more amazing native AI applications.

Li Yanhong said that Baidu's new search has three characteristics: extreme satisfaction, recommendation stimulation and multi-round interaction. when users search for questions, the new search will "no longer give you a bunch of links", but through the understanding of the content. generate multimodal answers for text, pictures and dynamic charts, allowing users to get answers in one step. In view of complex needs, the characteristics of "multi-round interaction" can also be prompted, adjusted and other ways to meet users' more personalized search needs.

At the same time, Li Yanhong also demonstrated the first domestic generative business intelligence product created with AI native thinking: Baidu GBI. According to reports, compared with the high threshold and difficult data analysis of traditional BI software, Baidu GBI can perform data query and analysis tasks through natural language interaction, and also support professional knowledge injection to meet more complex and professional analysis needs.

Through the understanding and regeneration of massive documents, pictures and videos, Baidu network disk and library have creative ability: the network disk can not only accurately locate a video frame, but also summarize an hour-long video content in a few seconds, and extract golden sentences and key points from it; the library is based on 1 billion high-quality materials, can achieve writing and PPT and other work, and become a veritable "productivity tool".

Baidu Maps and intelligent office platform are fluent, and through the ability of understanding and memory, they have become more intimate travel guides and super assistants: on maps, users only need to say their needs, and the map can transfer thousands of service interfaces to help users recommend restaurants, compare multi-location information, and give travel advice. For example, the flow can focus on the office pain points with a lot of group chat information, and the travel assistant can not only book air tickets and hotels, but also summarize the background information and conversation references for visiting customers by accessing company systems such as CRM.

As Robin Li said before, the native application of AI is not a simple repetition of App and PC software on the mobile Internet, but to be able to "solve problems that cannot be solved or can not be solved in the past."

Plug-ins and API help ecological prosperity and promote economic growth

"the big model will open up a prosperous AI native application ecology," Li stressed. Plug-ins are a special AI native application with the lowest threshold and the easiest to use, allowing developers and entrepreneurs to quickly join the ecology. He said, for example, that the "intelligent legal assistant" of the big model, which accesses authoritative legal data, can provide users with legal advice, while the resume assistant plug-in can help users generate resume templates with one click.

According to reports, the data, capabilities or applications of individuals and enterprises can quickly become AI plug-ins to enhance the ability of large models and make them more practical and easy to use. Li Yanhong said that a month ago, Baidu launched the Lingjing plug-in platform, and so far 27000 developers have applied to settle in, covering legal, workplace, learning and other fields.

When developing AI native applications, the basic ability of large models is very important. Li Yanhong said that API is the main way for AI native applications to call the basic large model. Enterprises and developers can access the large model API, including Wen Xin Yiyan, on Baidu's Qianfan big model platform. At present, Qianfan big model platform has become the largest large model development platform in China, with 42 mainstream large models settled in, covering nearly 500 scenes in various industries. From now on, enterprise customers can also apply to test the API of Wenxin 4.0 on the Qianfan model platform.

"China has rich application scenarios, and Chinese users are naturally willing to embrace new technologies. With advanced basic models, we can build a prosperous AI ecology and jointly create a new round of economic growth." Li Yanhong said.

In addition, Li Yanhong said that the future AI native applications must be multimodal, outside the information world, will certainly reconstruct the physical world. Autopilot is a typical application of large visual model to reconstruct the physical world. The big model will make Baidu's self-driving ability surpass the experience system, deal with complex scenes more intelligently, and achieve wider time and space coverage. At present, Baidu self-driving travel service platform radish fast run has provided more than 4 million services, making it the largest self-driving travel service provider in the world.

"A large number of native AI applications will continue to emerge, and digital technology and the real economy will be deeply integrated. Large models are becoming an important driving force for a new type of industrialization." Li Yanhong said. Just as the theme of Baidu World 2023 is "generating the Future", at the end of his speech, Robin Li announced that we are about to enter an era of native AI, an era of human-computer interaction through Prompt.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.