"In the future, natural language will become the new general-purpose programming language, and as long as you can speak, you can become a developer and change the world with your creativity. ”
On April 16, Create 2024 Baidu AI Developer Conference was held in Shenzhen. Robin Li, Founder, Chairman and CEO of Baidu, delivered a keynote speech entitled "Everyone is a Developer". He believes that large models and generative AI will revolutionize the developer community.
"AI is setting off a creative revolution, and developing applications in the future is as simple as shooting a short video, everyone is a developer, and everyone is a creator. ”
Baidu has prepared three "out-of-the-box" tools for developers, including AgentBuilder, an agent-based development tool, AppBuilder, an AI-native application development tool, and ModelBuilder, a model customization tool of various sizes. "These three tools all represent advanced productivity. ”
It is worth mentioning that Robin Li shared Baidu's specific ideas on the development of AI native applications on the spot, and said: "This is what our Baidu has stepped on countless pits and paid high tuition fees in exchange for the practice of the past year. "The three ideas for developing AI-native applications are MoE, small models, and agents.
At the meeting, Robin Li officially released the tool version of Wenxin Model 4.0. He also revealed that up to now, the number of users of Wenxin Yiyan has exceeded 200 million. "The Wenxin model has become the most advanced and widely used AI basic model in China. ”
Thanks to the power of the Wenxin large model, the smaller size model that developers cut out through Wenxin 4.0 dimensionality reduction is significantly better than the model directly called up from the open source model, and the cost is significantly lower under the same effect. "People used to use open source and thought that open source was cheap, but in fact, in the large model scenario, open source is the most expensive. So the open source model will fall further and further behind. ”
Here are the main points of Robin Li:
Natural language will become the new general-purpose programming language, and anyone who can speak can become a developer
"Big models and generative AI are going to revolutionize the developer community. In the past, developers used code to change the world, but in the future, natural language will become the new general-purpose programming language, and as long as you can speak, you can become a developer and change the world with your creativity. ”
"After more than a year on the job, Comate has entered tens of thousands of companies such as Himalaya, Mitsubishi Elevator, iSoftStone, etc., and the code adoption rate has reached 46%, and 27% of Baidu's new code every day has been generated by Comate. ”
"Today, you can make an AI application without coding, and you can make an agent without programming. AI is setting off a revolution in creativity, and developing applications in the future is as simple as shooting a short video, everyone is a developer, and everyone is a creator. ”
AgentBuilder、AppBuilder、ModelBuilder,都代表了先进生产力
"As a technology company, Baidu's role is to provide everyone with the development tools they need as much as possible to continuously improve the creativity of society as a whole. ”
"Specifically, we provide a powerful basic model series, which is the Wenxin large model series, which includes the flagship version of ERNIE3.5, ERNIE4.0, as well as the lightweight version of ERNIE Speed, Lite, Tiny, etc. ”
"We also provide tools to develop various applications based on large models, including AgentBuilder, an AI native application development tool, and ModelBuilder, a model customization tool of various sizes. These three tools all represent advanced productive forces. ”
The number of Wenxin Yiyan users exceeded 200 million, and the Wenxin model 4.0 tool version was officially released
"Wenxin Yiyan was released on March 16 last year, and it has been a year and a month since today. The number of our users has exceeded 200 million, the average daily API call volume has also exceeded 200 million, the number of customers served has reached 85,000, and the number of AI native applications developed using the Qianfan platform has exceeded 190,000. ”
"In recent months, Wenxin's large model has achieved further significant improvements in general capabilities such as code generation, code interpretation, and code optimization, reaching the international leading level. Today, we officially release the tool version of Wenxin Model 4.0. ”
"The Wenxin model has become the most advanced and widely used AI basic model in China. ”
"Compared with a year ago, the algorithm training efficiency of Wenxin's large model has been increased to 5.1 times, the average weekly training efficiency has reached 98.8%, the inference performance has been improved by 105 times, and the cost of inference has been reduced to 1%. In other words, customers can now make 1 million calls a day at the same cost instead of 10,000,000 calls a day. ”
The specific idea of developing AI-native applications is the result of stepping on countless pitfalls and paying high tuition fees in the past year
"Large models themselves do not directly create value, and AI applications developed based on large models can meet real market needs. ”
"What I would like to share with you today are some specific ideas and tools for developing AI native applications based on large models. This is what our Baidu has done in exchange for stepping on countless pits and paying high tuition fees according to the practice of the past year. ”
"The first is MoE. In the future, large-scale AI native applications will basically be MoE architecture, and the MoE mentioned here is not a general academic concept, but a mixture of large and small models, and does not rely on one model to solve all problems. ”
"The second is the small model. The inference cost of the small model is low, the response speed is fast, and in some specific scenarios, the use effect of the small model after SFT fine-tuning can be comparable to that of the large model. That's why we're releasing three lightweight models: Speed, Lite, and Tiny. We use a large model, compress and distill a basic model, and then use the data to train, which is much better than training a small model from scratch, and the effect is better, faster and lower than the model trained based on the open source model. ”
"The third is the agent. Agents are a hot topic at the moment, and with the improvement of their capabilities, a large number of new applications will continue to be born. The agent mechanism, which includes understanding, planning, reflection, and evolution, allows machines to think and act like humans, to complete complex tasks autonomously, and to continuously learn, iterate, and evolve in the environment. In some complex systems, we can also allow different agents to interact and cooperate with each other to complete tasks with higher quality. ”
Agents are the closest and most mainstream way to use large models for everyone in the future
"Agents may be the closest to everyone in the future, the most mainstream way to use large models, based on a powerful basic model, agents can be generated in batches and applied in a variety of scenarios. ”
"Baidu has just upgraded the Wenxin intelligent twin platform. Up to now, more than 30,000 agents have been created, more than 50,000 developers and tens of thousands of enterprises have settled in. Our goal is to let everyone and every organization become an agent developer, and create the most complete agent ecosystem in China. So how do you achieve this goal? It is to provide you with AgentBuilder, a zero-threshold agent development tool. ”
"Today, every merchant and every customer can have their own intelligent twins in Baidu. The whole process does not require programming at all, through the input of information similar to prompt words, and simple operation and optimization, you can quickly generate an agent and become a 7X24 hours online gold salesman. ”
At the scene, Robin Li demonstrated the three cases of three agents, Singapore Tourism Board, Kai Tak Education and Sophia, and taught developers to create an agent in 5 minutes and zero threshold in natural language.
"In the first week of its launch, it was successfully distributed 1.55 million times and interacted with users 58,000 times, with a linear increase in the number of lead conversions, a significant reduction in the conversion cost of effective leads, and a significant improvement in operational efficiency. ”
"Since the launch of Sophia's merchant agent, the cost of effective leads has dropped by 30%. That is, it gets a valid customer, if the cost used to be 100 yuan, now it only needs 70 yuan.
AppBuilder: The best AI-native application development tool, you can develop an application in three steps using natural language
"AppBuilder, it's the best AI-native app development tool out there. On AppBuilder, we have encapsulated and preset various components and frameworks required for developing AI-native applications in advance, greatly reducing the development threshold. ”
"In just three steps, developers can develop an AI-native application in natural language, which can be easily published and integrated into a variety of business environments. ”
At the scene, through three cases: "Playground Queuing Assistant", "Huadian AI Assistant" of North China Electric Power University, and Baidu Wenku Intelligent Comics Generation, Robin Li demonstrated the creation process of an AI native application. You only need to set the name, fill in the role instructions, and insert the component to create an AI-native application.
He also points out that AppBuilder has two major advantages:
"One is powerful. Relying on Wenxin 4.0's ability to understand and follow instructions, our AppBuilder can ensure that cold start-up can reach a good level, and will not take a long time to optimize due to poor results, which greatly reduces the development threshold. Relying on the retrieval enhancement technology RAG, in typical scenarios such as knowledge Q&A, our Q&A accuracy and friendly response degree have reached more than 95%, which greatly exceeds other similar products. ”
AppBuilder also provides a wealth of complete component tools, including 55 components such as Baidu Search based on Baidu's years of technology accumulation, AI capability components, large model capability components, and Baidu's exclusive open business components. and third-party APIs in some mainstream scenarios, such as flight query, paper query, etc. We've also just added support for custom components, so customers can connect directly to any of their proprietary tools and data. These rich components together support the efficient development of AI-native applications. ”
"Second, it's easy to use. With AppBuilder, you can quickly create and distribute apps in just three steps. We also support open-source SDKs, which is convenient for you to carry out secondary development. ”
ModelBuilder, a model customization tool for all sizes: Produce models efficiently and at low cost
"A more suitable tool for professional developers is ModelBuilder, which can customize models of any size according to the needs of developers, and further fine-tune the SFT according to the subdivision of the model, so as to achieve better results. ”
After data processing and model fine-tuning, the "Essay Correction Assistant" can not only have more professional teacher comment thinking and follow the format, but also compare with the unfine-tuned model, the fine-tuned model score is closer to the real teacher's comment score.
He also interacted with Xiaodu in real time on the spot, showing that Xiaodu uses the combination of multiple models of MoE to perform different tasks, such as using the small model ERNIE Tiny to perform model routing work, and Wenxin 4.0, which has the best performance, is used to perform complex requirements such as scheduling. According to reports,Compared with the flagship version that uses all Wenxin large models,Xiaodu can achieve a response speed increase of 2 times,Cost reduction99。
Robin Li said, "These examples of ModelBuilder show Baidu's ability to produce models efficiently and at low prices."
"In order to make it easier for everyone to get started quickly, ModelBuilder presets the most comprehensive and abundant large models. It includes ERNIE3.5 and ERNIE4.0, which are the flagship large models, which are suitable for general complex scenarios and have powerful capabilities, as well as three lightweight large models, ERNIE Speed, Lite, and Tiny, and two vertical scene models, ERNIE Character is suitable for role-playing, and ERNIE Functions is suitable for external tool use and business function calls in dialogue or Q&A scenarios. Of course, ModelBuilder also supports third-party mainstream models at home and abroad, with a total of 77 models, making it the development platform with the largest number of large models in China. ”
Open source models will fall further and further behind
"Because of Wenxin 4.0, the most powerful basic model, we can tailor a smaller model suitable for various scenarios according to our needs, taking into account various considerations such as effect, response speed, and inference cost, and support fine tuning and post pretrain. ”
"In this way, the model cut out through dimensionality reduction is significantly better than the model directly called up from the open source model, and the cost is significantly lower under the same size, and the cost is significantly lower under the same effect. ”
"People used to use open source and thought that open source was cheap, but in fact, in the large model scenario, open source is the most expensive. So the open source model will fall further and further behind. ”
Multi-modal large models are the only way to AGI, and the biggest application scenario of visual large models is autonomous driving
"Looking to the future, I believe that multimodal large models, or the integration of text, images, voice, video, etc., is a very important long-term development direction for basic models, and it is the only way to AGI. Baidu has a long-term investment in these areas and will keep up to date with the latest developments in large models. ”
"I have a very different judgment: the biggest application scenario of the visual model is autonomous driving. Baidu is the best in this direction, a global leader in autonomous driving, and we train AI not only on how to generate video, but also on AI to understand what's happening in the real world and predict the future. ”
"Based on more than 100 million kilometers of road test mileage data in complex cities in China, Baidu has trained the Apollo visual perception model. It has four basic capabilities: detection, tracking, understanding, and mapping. This gives Baidu a smarter, more adaptable, and safer autonomous driving solution. ”
Everyone can become a developer, and the future will be a future created by developers together
"Today's China, with 1 billion Internet users, a strong basic model, enough AI application scenarios, and the world's most complete industrial system, the state is also vigorously encouraging and supporting the 'artificial intelligence +' action, and every person and every enterprise only needs to make full use of these tools to unleash unlimited creativity and productivity. ”
"Everyone can be a developer, and the future will be a future created by developers together!"