"The emergence of intelligence brought by the big model is the basis for us to develop AI native applications." On October 17, Robin Lee said at Baidu World 2023. At the event site, Robin Lee made a speech with the theme of "Hand in Hand Teaching You to Make AI Native Applications", released the 4.0 version of Wenxin large models, and brought more than 10 AI native applications such as new search and new maps.
At the conference, Robin Lee announced that Wenxin large models 4.0 was officially released and invited to test. He said that this is the most powerful Wenxin large models so far, which has realized the comprehensive upgrade of the basic model, and has significantly improved the ability of understanding, generation, logic and memory. The comprehensive ability is "no inferior to GPT-4". Robin Lee introduced that Wenxin 4.0 was also invited to test at the same time. The on-site audience could scan the QR code of the guest card, log on to the ERNIE Bot official website or download the latest version of ERNIE Bot APP to experience it; In addition, enterprise customers can also apply for testing Wenxin 4.0 API through Baidu AI Cloud Qianfan model platform.
Robin Lee showed more than 10 AI native applications based on ERNIE Bot's reconstruction, such as Baidu Search, Ruliu, Map, Netdisk, Library, and so on. He hoped to expand everyone's imagination and "inspire everyone to make more amazing AI native applications together".
The strongest Wenxin large models 4.0 release
In Robin Lee's view, the birth of AI native applications benefited from the four core capabilities of understanding, generation, logic and memory of the big model. Baidu's AI native applications were also developed based on ERNIE Bot's words.
Based on Wenxin large models 4.0, Robin Lee demonstrated the characteristics and application scenarios of the four capabilities in turn. In terms of understanding ability, he showed ERNIE Bot's understanding of complex prompts such as disorderly order, vague intention and subtext by asking about the case of the provident fund's off-site loan policy.
In terms of generating ability, Robin Lee showed how ERNIE Bot quickly generated a group of advertising posters, five advertising copies and a marketing video according to a picture of the material in just a few minutes. It is reported that based on this series of capabilities, Baidu has launched the AIGC marketing creative platform Qingrud, allowing "one person to become an AI marketing team".
At the same time, he also demonstrated the logical ability of large models through scenarios such as solving mathematical problems and summarizing knowledge points; Through thousands of words of novel writing and character and plot settings, it embodies the memory ability of the big model; And digital doctors help patients interpret drug instructions to demonstrate the comprehensive application of the four abilities.
"The previous demonstration reflects the progress of the Wenxin large models in understanding, generation, logic and memory, which are the basis for the survival of all AI native applications." Robin Lee said.
More than ten AI native applications have been heavily released
Rich AI native applications are the value of big models. At the conference, Robin Lee announced that "our search, Ruliu, maps, online disks, libraries, etc. will meet with you in a new face". He stated that the purpose of sharing these applications is to expand imagination and inspire more people to create more stunning AI native applications.
Robin Lee introduced that Baidu's new search has three characteristics: extreme satisfaction, recommendation stimulation and multiple rounds of interaction. When users search for questions, the new search will "no longer give you a bunch of links", but generate multimodal answers of text, pictures and dynamic charts through understanding the content, so that users can get answers in one step. When dealing with complex needs, the "multi round interaction" feature can also meet users' more personalized search needs through prompts, adjustments, and other methods.
At the same time, Robin Lee also showed Baidu GBI, the first generative business intelligence product in China built with AI native thinking. It is reported that compared to the high threshold and difficulty in 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.
By understanding and regenerating massive amounts of documents, images, and videos, Baidu Netdisk and Wenku have gained creative capabilities. Netdisk can not only accurately locate a certain frame of the video, but also summarize the video content for up to one hour in a few seconds, and extract golden sentences and key points from it; The library is based on 1 billion high-quality materials, which can achieve tasks such as writing and PPT, becoming a truly "productivity tool".
Baidu Maps and intelligent office platforms are like a stream, and through their abilities such as understanding and memory, they have become more thoughtful travel guides and super assistants. On the map, users only need to express their needs, and the map can mobilize thousands of service interfaces to help users recommend restaurants, compare information from multiple locations, and provide travel suggestions; Ruliu can focus on the office pain points with a lot of group chat information in one second. The travel assistant can not only book airline tickets and hotels, but also summarize the background information and conversation reference of visiting customers by connecting to company systems such as CRM.
As Robin Lee said earlier, AI native applications are not simple repetitions of mobile Internet apps and PC software, but should be able to "solve problems that could not or could not be solved in the past".
Plugins and APIs help promote ecological prosperity and economic growth
"The big model will open a prosperous AI native application ecosystem." Robin Lee stressed that plug-in is a special AI native application, with the lowest threshold and the easiest way to get started, which enables developers and entrepreneurs to quickly join the ecosystem. For example, a "smart legal assistant" that connects big models to authoritative legal data can provide users with legal advice, while a resume assistant plugin can help users generate resume templates with just one click.
It is reported that personal and corporate data, capabilities, or applications can quickly become AI plugins, enhancing the capabilities of large models and making them more practical and easy to use. Robin Lee said that one month ago, Baidu launched the Lingjing plug-in platform. At present, 27000 developers have applied to settle in, covering many fields such as law, workplace, and learning.
When developing AI native applications, the basic capabilities of large models are crucial. Robin Lee introduced that API is the main way for AI native applications to call the basic big model. Enterprises and developers can access the big model API including ERNIE Bot on Baidu's Qianfan big model platform. At present, Qianfan big model platform has become the largest big model development platform in China, with 42 mainstream big models settled, covering nearly 500 scenes in all walks of life. Starting from now on, corporate 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 an advanced basic model, we can build a prosperous AI ecosystem and jointly create a new round of economic growth." Robin Lee said.
In addition, Robin Lee said that the future AI native applications must be multi-modal, and the physical world will be reconstructed beyond the information world. Autonomous driving is a typical application of reconstructing the physical world from visual models. The large model will enable Baidu's autonomous driving ability to surpass the experience system, handle complex scenes more intelligently, and achieve a wider range of spatiotemporal coverage. At present, Baidu's autonomous driving travel service platform, Radish Run, has provided over 4 million services and has become the world's largest autonomous driving travel service provider.
"A large number of AI native applications will continue to emerge, and digital technology and the real economy will be deeply integrated... The big model is becoming an important driving force for new industrialization." Robin Lee said. As the theme of Baidu World 2023 is "generating the future", at the end of the speech, Robin Lee said that we are about to enter an era of AI nativeness, an era of human-computer interaction through Prompt.