From government, academia, business circles, investment circles to entrepreneurs, artificial intelligence is regarded as the future direction; The massive analysis reports of analysts and media practitioners have made artificial intelligence quickly occupy the audio-visual of every ordinary person.
Ever since, just like the rise of every industry in history, artificial intelligence has gradually become a bit "confusing and fascinating" while singing and dancing.
On the positive side, artificial intelligence has spawned a large number of new technologies, new enterprises and new formats, provided new economic growth points for individuals, enterprises, countries and even the world, and even led to the fourth technological revolution, creating great value.
IDC predicts that the global artificial intelligence expenditure will reach 275.8 billion RMB by 2020, and the compound annual growth rate will exceed 50% in the next five years. The expenditure on artificial intelligence technology in China will reach 32.5 billion yuan, accounting for 12% of the total global expenditure.
On the negative side, although artificial intelligence has opened a brand-new era, it is also constantly breeding "bubbles", boasting about it, following the trend and making gimmicks. In the past two years, dozens of Chinese and American AI startups have closed down intensively, and a large number of AI startup projects have died halfway, which inevitably makes people feel that artificial intelligence just "looks beautiful".
So, what will the future of artificial intelligence develop into? How can we really land? How to realize the commercialization of scale? Although the concept of artificial intelligence has been put forward for more than 60 years, there are not many traces of predecessors in the combination of theory, technology, application and commerce.
Therefore, in the development of artificial intelligence industry, the roles of "pioneers" and "leaders" are particularly important.
"Scene-driven" is the key to AI landing
Among the many players of artificial intelligence, Alibaba has been trying to become this role. For the future of AI, Ali has a clear understanding and a unique AI development path.
12 on February 20th, Hu Xiaoming, president of Alibaba Cloud, put forward the concept of "industry-oriented AI": artificial intelligence is not only a "conceptual AI" in laboratories and PPT, but also an "industrial AI".
Hu Xiaoming said that the proposal of "industrial AI" is based on Alibaba's three judgments on artificial intelligence:
"First, there must be a scene-driven, what problems are we solving, how much the cost of this society has decreased, and how much the efficiency has improved; Second, is there enough data behind artificial intelligence to drive the improvement of AI capabilities? Third, whether there is enough computing power to support our algorithm and deep learning; Artificial intelligence will be valuable only if three scenarios are available at the same time. "
Ali puts "scene-driven" in the first place, which is the core of Ali's "industrial AI" strategy, Ali's unique AI development path and the exclusive secret of Ali's AI landing.
Different from the practice of many enterprises and institutions, Ali's AI journey does not start from the research and discussion in the laboratory, but runs in the opposite direction, starting from the basic business department and making AI "grow out" from daily scenes.
For example, the "Pat Li Tao" function in mobile phone Taobao allows users to search for pictures by taking photos, which is derived from e-commerce scenarios, and then solve technical problems one by one, and finally form a mature AI solution.
The e-commerce platform provides Ali with an excellent soil for AI growth. The universal or personalized needs of a large number of consumers have created different application scenarios; Massive data provides sufficient "raw materials" for AI; Alibaba Cloud's powerful computing power has become an accelerator for AI implementation. With these three elements, Ali can make artificial intelligence play its value quickly.
Facts have proved that Ali's industrial AI path of "bottom-up", "coming from the scene" and "going to the scene again" is correct and effective, which has promoted the real landing of AI technology in industrial application scenarios.
On the day of "Double 1 1", the robot customer service "Ali Xiaomi" undertook 95% of customer service consultation; The machine intelligence recommendation system has produced more than 56.7 billion exclusive shelves; AI designer "Luban" designed 4 1 10,000 product posters during the double 1 1 period; The operation and maintenance robot of Ali North China Data Center took over 30% repetitive work of operation and maintenance personnel.
Not only in the retail field, Ali's "industrial AI" layout has covered ten vertical fields such as city, finance, justice, agriculture, education, aviation, industry, security, environment and medical care, and has blossomed one after another, aiming at reshaping vertical industries with AI technology.
For example, in the financial field, Ali reduced the manual video verification in the process of applying for loans by 54% through cloud computing and intelligent algorithms; In the industrial field, Alibaba Cloud ET industrial brain helps Trina Solar to improve the battery A rate by 7%; In the field of smart cities, the brain of Alibaba Cloud ET City took over the traffic lights at the intersection of Hangzhou 128. Through the overall analysis of video and other data, the speed and efficiency of road operation were optimized, and the road transit time in the pilot area was reduced by 15.3%.
In Hu Xiaoming's view, every industrial revolution in the past was a deep integration of technology and industry, which led to economic and social changes, and AI was no exception. In the future, AI should go deep into all walks of life and solve the thorny problems encountered in life, production and social environment, thus leading the real industrial revolution.
Through the layout of "industrial AI", Ali gradually drifted away on the road of "deep integration of industry and AI".
"ET Brain" Let the Industry Share the AI Dividend
A year ago, Alibaba Cloud released artificial intelligence ET, which fully integrated Alibaba's voice, image, face, natural language understanding and other capabilities. At the Yun Qi Conference-Beijing Summit on February 20th, 65438, Alibaba Cloud officially launched the ET brain, which integrates global capabilities such as urban management, industrial optimization, auxiliary medical care, environmental management and aviation scheduling, upgrading the skills of ET from a single point to an ET brain with global intelligence, and comprehensively laying out industrial AI.
ET brain marker
According to Min Wanli, the chief scientist of machine intelligence in Alibaba Cloud, the core competence of ET brain is "quantum topology", and its birth can be traced back to Einstein's paper on Brownian motion published in 1905: "How to find its core path from these propagation appearances in a huge network? This is precisely the core ability of ET brain and a unique ability. "
Min Wanli said that compared with other AI products, Alibaba Cloud ET Brain combined AI technology, cloud computing big data capability with vertical industry knowledge, and realized the technological leap from single point intelligence to multi-agent intelligence based on the physical architecture of brain-like neural network and fuzzy cognitive inversion theory, creating a super agent with brain-like cognitive abilities such as multidimensional perception, global insight, real-time decision-making and continuous evolution.
The release of ET Brain means that Alibaba Cloud's AI capability has evolved from a single-point technology to a global capability for vertical industries. In the past year, ET brain has gained a lot of practice in cities, industries, medical care and other fields. Quantitative change leads to qualitative change, and then it can be upgraded to the "brain" of various industries. Min Wanli said that the ET brain will be set as an open ecology, allowing startups, developers and industry companies to share the dividends of technology.
In addition to ET Brain, Alibaba Cloud also released ET Aviation Brain at Yun Qi Conference and Beijing Summit, using artificial intelligence methods such as operation optimization and machine learning to allocate parking spaces. It is estimated that 1700 flights will be dispatched every day, which will help passengers save 5000 hours, greatly improve the flight transfer efficiency, and thus reduce the delay rate.
According to Li, the intelligent scheduling of airport stands is only one of the functions of ET aviation brain, and it also hopes to go deep into other aviation scenes. Previously, Alibaba Cloud Tianchi Platform, Xiamen Airlines and Baiyun Airport jointly launched the Smart Aviation AI Competition, inviting global engineers to use intelligent algorithms to solve problems in aviation scenes. In the future, ET Aviation Brain will continue to provide artificial intelligence solutions for flight intelligent recovery, airport ground scheduling and route planning. To build smart aviation.
At the Yun Qi Conference and Beijing Summit, Alibaba Cloud also announced the launch of an intelligent decision-making financial scheme-ET Financial Brain, which has the capabilities of intelligent risk control, thousands of people, networking and intelligent customer service.
According to Min Xu, general manager of Alibaba Cloud Finance Department, ET financial brain can assist banks, securities, insurance and other financial institutions to realize intelligent decision-making and risk control supervision on loans, credit information and insurance, greatly reduce the asset loss rate, improve the forecast accuracy of credit cards, and promote financial institutions to explore inclusive finance aspects such as Internet consumer finance and financial services for small and medium-sized enterprises.
Nowadays, ET financial brain has been applied in Nanjing Bank, Zheshang Bank, Guangfa Bank and other financial institutions. , and showed his talents in intelligent risk control, "thousands of people and thousands of faces" financial services, and the development of "new finance" business models.
Push the land to promote ecology and let AI "benefit the public"
From Terminator, The Matrix to westworld, human beings have expressed their worries about artificial intelligence. Will artificial intelligence replace human beings in the future? Alex Pentland, director of the Human Dynamics Laboratory at the Massachusetts Institute of Technology and author of "Smart Society", once pointed out that what we should worry about is not the global artificial intelligence itself, but its behind-the-scenes manipulator.
Artificial intelligence is a tool created by human beings. Now the industry is more willing to position artificial intelligence as "enhanced intelligence". Its goal is not to replace human beings, but to enhance human capabilities and serve human production and life. Therefore, artificial intelligence should not be closed in the laboratory, but should be closely integrated with human production and life to benefit the public.
The "industrial AI" advocated by Ali is just a way to open its AI capabilities and benefit the public. Ali AI capabilities have successively landed in cities, industries, automobiles, retail, finance, home, aviation and other fields. While solving industry problems, it has also brought changes to the lives of ordinary consumers, allowing consumers to effectively benefit from AI.
Ali is also making AI capabilities benefit more industries and consumers through open cooperation.
At the Yun Qi Conference and the Beijing Summit, Alibaba Cloud and China Telecom cooperated in the security field, and the two sides will jointly launch customized DDoS protection services next year to provide inclusive security for SMEs; Xinhua Bookstore cooperated with Alibaba Cloud to set up a smart bookstore, establishing close contact with customers on the consumer side, and creating a brand-new concept of "reading life" to meet the diversified and personalized needs of consumers.
At the same time, Alibaba Cloud announced that it has reached a strategic cooperation with Longping Hi-Tech and CITIC Cloud, and plans to push ET brain into the agricultural field, which is mainly used for screening and breeding, digitalization of infrastructure, agricultural management, site selection and crop yield prediction. Alibaba Cloud and BMW China officially announced that they will provide BMW owners with one-stop seamless remote service from home to car based on the Internet of Things, so as to query the real-time status of cars and remotely control their smart life.
In addition to deeply combining AI technology with the industry to benefit the general public, Ali is also actively participating in the construction of artificial intelligence ecology and the cultivation of talents, and promoting the accelerated landing of artificial intelligence in China.
At the Yun Qi Conference and the Beijing Summit, Alibaba Cloud, together with Zhang Tong Jiayuan, Becha, Wisdom Tree, Ant Technology, Detu and other manufacturers, released the "AI Vision Guardian Alliance", hoping to deeply integrate artificial intelligence and video technology with industries such as industry, agriculture and education.
Alibaba Cloud is deeply integrated into the national big data strategy, including deeply participating in two national big data engineering laboratories initiated by the government; Alibaba Cloud Tianchi Big Data Platform has gathered more than 1 10000 developers; Ali also launched the Pratt & Whitney Program with 307 universities in China, bringing new technologies such as cloud computing, big data and artificial intelligence into universities and cultivating talents.
To sum up, the development and landing of artificial intelligence in China need not only "pioneers" and "leaders", but also the prosperity and healthy development of the whole ecology. In this process, Ali and his "industrial AI" played a key role.