Shulou(Shulou.com)06/02 Report--

Recently, there is such a new discovery: originally talking to some friends outside the AI industry about AI, people pay more attention to algorithms and consumer hardware. However, when talking about this topic recently, everyone's attention has come to the issue of how to make good use of AI in various industries, which is relatively farther away from public life and more professional.

I think this is because today's media, data institutions and academic research achievements are all paying attention to such a logic: 90% of the development space of AI in the future is to be combined with the industry.

Whether it is called industrial AI, industrial Internet, or AI to B, and the "smart plus" mentioned in this year's "government work report", there is the same essence behind various references: AI technology must be combined with thousands of industries and a variety of industrial entities, and be integrated into the production core areas of the industry like a lubricant, so as to stimulate production efficiency and improve quality and efficiency for the industry.

However, how to put this big judgment into the details is a relatively vague thing for countless practitioners in today's industry. After all, AI technology is a new proposition, and the knowledge system and industrial cognition of different industries are very different. It is the common expectation of today's AI industry and social economy that how to effectively and accurately answer the question of how to integrate AI and enterprise market.

On April 9, at the graduation ceremony of the first students of Alpha College of Baidu University, Tsinghua University and Baidu jointly issued the "White Paper on Industrial Intelligence". This is the first industry report in China to discuss the process and current situation of the integration of AI and industry from the perspective of industrial evolution based on AI practice. It is also Baidu Intelligent Cloud after helping many enterprises successfully achieve intelligent upgrading. For the first time, it publicly expounds the strategic thinking and business practice methodology of industrial intelligence.

From this "white paper", we can see from several new horizons that Baidu is giving unique answers to several key questions: the proposition of intelligence + industry, why on earth do it? How do you do that? Who will do it?

Let's go layer by layer to find answers to these three questions and take a peek at the actual situation of the intelligent development of China's industry that the White Paper shows us.

Why do you want to do industrial intelligence? Because it's the meeting point of five requirements.

Through the introduction of AI technology and AI industry, as well as the landing of AI technology in many industries, the White Paper deduces the role of AI in promoting social economy and the strategic position of China's AI in global development. With detailed data, we can see that AI is on the road to industry, which happens to be in the intersection of several major needs from technology, industry, and then to the national economy.

1. From the perspective of technical logic, the basic characteristic of AI technology is to change the way of interaction and reorganize the value of data, which requires this technology to achieve value in the penetration of the industry and improve social production efficiency through the transformation of a large number of production details.

2. AI industry demand: at present, the AI technology accumulated by the science and technology industry needs to be released. Having a wide range of needs, specific scenarios, a large number of structured industrial markets, industry markets are the best goals, which can maximize the value of AI technology and the initiative of AI enterprises. Baidu Smart Cloud, for example, can use Baidu's underlying AI technology to achieve effective output and industry solution construction through B-end market cooperation. Let Baidu's AI technology first-mover advantage and leading layout become the driving force of technological innovation in the industry.

3. From the point of view of B-end market demand, the economy has entered a new normal, whether it is the real economy or the Internet economy, are entering a new stage. In this case, AI acts directly on the production cycle, which is in line with the core demand of using technology to drive the development of original productive forces, and achieve the improvement of quality and efficiency widely desired by enterprises. Therefore, the logical demand for "intelligence +" is a more profound market expansion than the previous enterprise + technology market.

4. From the perspective of social and economic needs, the deep combination of AI technology and industry can optimize the allocation of labor force, re-stimulate talent dividends, and continue to promote entrepreneurship, which has a wide range of social value.

5. From the perspective of national strategy, "intelligence + industry" is based on China's unique exploration of AI, as well as the unique law of national economy and industrial chain. Today, many international media are lamenting that China's AI industry will enter the unknown areas of the world of human technology. This is a technological leap led by China, and to some extent even the first scientific and technological exploration guided by non-Western civilization. Therefore, the wide application of industrial intelligence can be seen as an attempt to build a community with a shared future for mankind, and an important opportunity for China to lead technological exploration and feed global technological innovation.

Therefore, the convergence of technology, the joint efforts of social enterprises and the integration of industry and learning to explore industrial intelligence strategy is a profound attempt of technology application in China's social economy today. It is clear that there is a good reason for industrial intelligence to do things, and then we will face the specific problem of how to integrate AI with the industry.

How can enterprises move towards the "smart +" era? the White Paper gives suggestions in four directions.

AI really enters the industrial side, can not rely on empty talk, or in accordance with the general strategic direction, but must proceed from the pragmatic and rational logic, analyze the stages of industrial intelligence, and give the idea of gradual development.

This is Baidu's strategic way of thinking that we found in the White Paper. Different from the generally vague cognition of the AI toB industry, Baidu gives a more pragmatic and accurate methodological thinking on the basis of a large number of AI technology research and industrial practice.

In the "White Paper", it objectively analyzes the development of today's AI technology and the acceptance of the industry. Standing in Baidu's AI technology research and evaluation system and cooperation case, discuss how to implement industrial intelligence in the end. It is pointed out in the "White Paper" that the laws of four directions are the weather vane of today's AI moving towards industrial application:

1. Enterprises need to integrate into AI in details through digital structure.

Not every industry can be casually and comprehensively AI. It is pointed out in the "White Paper" that the industrialization process of AI today is the exploration and application of many different AI in different industries at the same time. Then the grasp of vertical technology and vertical scenarios, as well as the informationization and standardization of enterprise business processes, and the digitization and structure of business data are important prerequisites.

In a word, the premise of developing AI industrialization is the digital construction of the enterprise itself and the objective understanding of the integration mode of AI technology system and details.

2. The industrialization of several hot technologies has a higher degree of maturity and can be infiltrated first.

Not every AI technology can be applied industrially on a large scale. The White Paper objectively points out that today, the industrialization maturity of several hot technologies such as intelligent voice, machine vision, machine recommendation, and so on is relatively higher. The industrial application of AI should also take the lead from these technologies. We should not blindly pursue the cutting edge and large-scale application of AI technology.

3. Establish an extensive and in-depth industrial chain to achieve an effective connection among computing layer, algorithm layer, application layer and user layer.

AI can not be built by a company. It must be closely integrated with the industrial chain to establish an ecological organization based on AI industrialization. According to the White Paper, the main task of today's industry chain is to establish a two-way relationship between AI application layer manufacturers and AI algorithm layer manufacturers, and to communicate industry needs and technical directions acutely. This is also Baidu's main layout idea of ecological construction based on to B business today.

4. The four technologies are organized together to become the infrastructure, and the leading enterprises of AI serve as the driving force for development.

AI is not a castle in the air, nor can it fight alone to become a practical AI. It must be based on the conditions that together with big data, cloud and the Internet of things to form an industrial infrastructure. Their coordinated development can promote the final realization of the new paradigm of industrial intelligence.

On the basis of this logic, large science and technology enterprises with AI algorithm technology, big data and cloud technology must undertake more comprehensive and profound tasks to promote the development of industrialization. For example, Baidu Smart Cloud plays the role of such industry driving force in the industrial layout and acceleration stage of ABC+IoT, actively exploring a large number of unknown areas of industrial intelligence.

After understanding how the integration of AI and industry hits the ground, the last and most important issue is the role sharing and coordinated development in this big chess game. According to the "White Paper", the coordination of the three industrial roles is the power engine for industrial intelligence to move towards the flashpoint.

Who will promote development? Industrial intelligence is heading towards the flashpoint, and each of the three roles must perform their respective duties.

The industrial intelligence system with effective promotion and ecological acceleration is based on the coordinated development of technology providers, industrial application transformation layer and industry audience. Among them, the artificial intelligence enterprises represented by Baidu can be said to be the basic driving force to accelerate the promotion of industrial intelligence today.

In the logic of the entire industry intelligence industry, the three roles must perform their respective duties:

1. The giant AI, as a "heavy weapon of the country", must assume four roles.

After extensive technology accumulation, industrial output system construction and industrial intelligence practice, AI leaders like Baidu have become a veritable driving force for the development of industrial AI and an important national tool for the exploration of AI industrialization in China today. The role of the leader must be more complex and diverse. For example, Baidu is increasingly playing four roles in the development path of industrial intelligence:

Role one: technology engine, as Baidu's investment layout in the field of AI is unparalleled, its value as the technology source of industrial AI is becoming more and more prominent.

Role 2: industrial exporter, since the adjustment of the structure, Baidu's strategic layout and business actions in AI toB have been significantly accelerated, and a complete and ecologically orderly industry infrastructure has been built with the strategy of AI+ Cloud. As the export of industrial exploration, Baidu's AI+ cloud system must undertake more practical tasks and contribute to the strategic position of weather vane and navigator.

Role 3: ecological explorer, another task of core AI enterprises is to establish an effective industrial cooperation map to prove the feasibility and long-term development value of industrial AI ecology. On this basis, Baidu has not only made remarkable achievements in promoting the industrial application of AI, completed the landing cases of AI in many industries, such as SAIC car booking, Pacific Insurance, China Unicom, Jimi Technology, Softcom Wisdom, etc., but also carried out extensive exploration in the fields of ecological empowerment and talent education, contributing multiple driving forces to industrial intelligence and ecology.

Role 4: regular thinker. This "white paper" not only shows the development status of Baidu in the integration of AI and industry, but also reveals its deep thinking on the industrial intelligent market. For example, on the basis of actual cases, the TUMC model is explored to help enterprises better understand the application ability of AI, and assume the responsibility of top technology companies at the strategic and knowledge leading level.

In the stage when industrial intelligence must develop at a high speed, AI leading enterprises must play the role of engine at many levels, and provide industry wisdom and development cases that can be used for reference. This is also the main task and contribution of Baidu at this stage.

2. The industry needs AI application companies to be the pioneers of industrial intelligence exploration.

In the AI application layer, there must be industry technology companies and AI application enterprises to shoulder the hub work of AI entering the industry, to seek industry knowledge, to explore application scenarios and replicable business models. At the same time, in the transition from technology to application, fully serve the personalized needs of AI in the B-end market, and create new social value and employment opportunities at the same time.

3. Active cooperation and rational sensitivity are the soil for AI to blossom.

Various industries as users need to actively accommodate new technologies and seek new solutions to improve productivity in this process of technological change. Welcome "Smart +" with open and inclusive, forward-looking strategic vision and technology investment.

Only when these three roles perform their respective duties and cooperate effectively, can industrial intelligence really approach the flashpoint. It should be noted that the exploration trajectory of industrial intelligence is a new scientific and technological proposition all over the world today. Today, China has not only the accumulation of AI technology that can influence the world, but also the industrial market with complex depth and multi-level demand. The combination of the two will be an unprecedented exploration in the history of science and technology application.

This is the opportunity of the times, the opportunity of Baidu and other AI enterprises, the opportunity of thousands of industries, but also the opportunity of China.

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.