Summary: The development of artificial intelligence will not stop here. The development of AI based on the Internet cloud-brain architecture will further develop the relationship between the various neural systems of the Internet, which will become a trend and inevitable. Cloud reflex arcs therefore Will become the next development focus of artificial intelligence applications. This article was transferred from | Artificial Intelligence çŠ é«¡ 刘 | Liu Feng Since 2014, artificial intelligence has gradually become the most popular concept in the field of science and technology, and has been widely followed by the scientific and technological circles, the business community and the media. As an academic field, artificial intelligence came together in the summer of 1956 with a group of visionary young scientists headed by McCarthy, Minsky, Rochester, and Shennong to discuss together and explore the use of machines to simulate intelligence. A series of related questions were first raised. In fact, the development of artificial intelligence is full of ups and downs. In the past 60 years, artificial intelligence has experienced many stages from optimism to pessimism and from high tide to low tide. The most recent downturn occurred in Japan's fifth-generation computer program in 1992 and ended without results. Subsequently, the artificial neural network fever began to feverish in the early 1990s, and the artificial intelligence field once again entered the "winter of AI." This winter was so cold and long, until Geoffrey Hinton, professor at the University of Toronto, Canada, proposed the "deep learning" algorithm in 2006, and the situation changed. This algorithm is an ingenious upgrade of the artificial neural network theory born in the 1940s. Its biggest innovation is the effective processing of huge data. This feature is fortunately combined with the Internet. This triggered a new wave of artificial intelligence since 2010. In 2011, an NCAP researcher and Stanford’s Andrew Ng established Google’s brain based on deep learning in Google, and Andrew Ng was later the chief scientist of Baidu’s brain, Wu Enda. In 2013, Geoffrey Hinton joined Google, whose purpose is to further deepen the work of Google's brain. Artificial intelligence has entered a new era from now on--Internet artificial intelligence era, massive Internet-based "big data" and information exchanged with the real world at any moment, including Amazon, Facebook, Baidu, Tencent, Alibaba, Microsoft, Intel Giants such as IBM have entered the AI ​​field. Continuously produce new results and create new records. It should be said that this round of artificial intelligence boom is essentially still another wave-like climax in the evolution of the Internet. It can not be separated from the Internet before the application and technology lay the foundation for the new outbreak of artificial intelligence. Moreover, from the perspective of AI leaders, it is mainly Internet giants, how artificial intelligence actually broke out, and what form it will follow in the future. This article will focus on discussion. I. The development of the Internet's various nervous systems lays the foundation for an artificial intelligence outbreak Since 2007, the relevant research team of the Academy of Sciences has published a paper proposing the future trend of the Internet: “The Internet is evolving in a direction that is highly similar to the human brain. It will have its own visual, auditory, tactile, motor nervous system and will have its own memory. The nervous system, the central nervous system, and the autonomic nervous system.". This resulted in the Internet Cloud Brain architecture (Reference 1). By 2017, with the rapid development of science and technology such as artificial intelligence, Internet of Things, big data, cloud computing, robotics, virtual reality, and industrial Internet, the structure of the internet cloud will gradually become clear. From 2008 to 2017, it will form 4 Version of the Internet Cloud Brain Architecture Diagram. It should be said that the internet cloud brain is a kind of brain-like structure formed during the evolution of the Internet. It is not mature and complete at first, but it has gradually developed during the process of scientific exploration and commercial innovation. Web2.0, Internet of Things, Cloud Computing, Industrial Internet, Big Data, Artificial Intelligence. . They are not new things that are not separated from the Internet, but are the wave-like peaks caused by the uneven development of various nervous systems during the development of the Internet's cloud and brain, often after a technology or pattern matures, the next technology or pattern flourishes. Foundation. We can look at the progress of the Internet before the artificial intelligence boom. 1. The Internet of Things is essentially the Internet's central nervous system and its controlled sensory and motor systems. Its development provides artificial intelligence with a large amount of sensor data. 2. Cloud computing is essentially the central nervous system of the Internet, which uses servers, network operating systems, neuron networks (large social networks), big data, and artificial intelligence algorithms based on big data to other components of the Internet's cloud and brain. Control. 3. Big data is essentially the valuable information transmitted and accumulated during the operation of the Internet's clouds and brains. Because of the rapid expansion of the Internet over the past 50 years, the volume has become enormous. The combination of big data and artificial intelligence algorithms is the basis for the Internet's cloud and brain to produce intelligent intelligence. 4.Industry 4.0 and Industrial Internet are essentially the motor and nervous system of the Internet, which will be a very large part of the future of the Internet cloud and brain. Its development will provide a wide range of data for automobiles, robots, office equipment, and production equipment for artificial intelligence. , and application scenarios. 5. Big Sns is an internet cloud brain neural network. It was developed from the Internet's traditional social networks Facebook, WeChat, and Weibo. It evolved from linking people to people, linking people and things, things and things, and even including link labor. Intelligent software system, its development has established a close relationship between artificial intelligence and physical intelligence. At the same time, artificial intelligence system can become a part of social network that can interact with other people or things. In summary, this round of artificial intelligence is essentially the source of power generated by the Internet's cloud and brain to produce intelligence intelligence. Artificial intelligence is not only combined through algorithms such as deep learning, machine learning and big data, but also applied to the nerve endings of the Internet's cloud and brain. , neural networks and smart terminals. Makes the Internet’s various brain systems enhance their capabilities at the same time. Second, the Internet + AI to promote the development of cloud reflex arc Currently, artificial intelligence technology is widely used in various components of the Internet: The combination of artificial intelligence and the Internet's central nervous system has produced Google's brain, Baidu's brain, Ali cloud, Amazon cloud, Tencent cloud and other cloud artificial intelligence giant systems; The combination of artificial intelligence and the auditory nervous system of the Internet generates new sound identification products such as HKUST's X-Ray and Yun Zhisheng. Artificial intelligence, especially deep learning combined with the Internet's visual-sensing nervous system, produces new image recognition products such as Green's deep pupil, Face++, etc. The combination of artificial intelligence and the Internet's motor nervous system has produced new applications such as smart manufacturing, smart driving, and cloud robots; The combination of artificial intelligence and internet neural networks (large social networks) has produced intelligent virtual assistant products such as Duwei and Xiaobing; The combination of artificial intelligence and the sensory nervous system of the Internet has brought about innovative applications of edge computing. However, the development of artificial intelligence will not stop there. The development of AI based on the Internet cloud-brain architecture will further develop the relationship between the various neural systems of the Internet, and it will become a trend and inevitable. Cloud reflex arcs will therefore Become the next development focus of artificial intelligence applications. Third, the nine different types and composition of cloud reflex arcs Neuroreflection is one of the most important neural activities in the human nervous system. It was first proposed by the French philosopher Descartes. He noticed that the body has a regular response to some environmental stimuli. For example, a foreign object hits the cornea and causes a blink. He borrowed the concept of reflection in physics, thinking that animal activity is reflected as light is cast onto a mirror. Analytical methods of studying the brain and spinal cord and clarifying the laws of reflection are Sherrington in the United Kingdom and Chechenov and Pavlov in Russia. The reflex arc is the structural basis of reflex activity. It is the entire path of the body's movement through the nervous system from the stimulation to the reaction. Reflections generally require a complete reflex arc. A complete reflex arc consists of five basic parts: sensor, afferent nerve, nerve center, efferent nerve, and effector. Knee-jump reflex is the most familiar and the simplest reflex activity. The Shou-Key Blowing, the Ae-Hui, the Ai-Fu, and the Ash-A-Ship, said that the cylinder, the hard, the round, the South, the porcelain, and E-mail playing ヌ 渚 桓霰曜嫉 桓霰曜嫉 桓霰曜嫉 桓霰曜嫉 桓霰曜嫉 桓霰曜嫉 桓霰曜嫉 桓霰曜嫉 桓霰曜嫉 桓霰曜嫉 æµ âˆ çª’èŸ† æµ âˆ çª’èŸ† æµ âˆ çª’èŸ† æµ âˆ çª’èŸ† æµ âˆ çª’èŸ† æµ âˆ çª’èŸ† Α Α Α Α Α Α 㶠㶠㶠㶠ã¶æŒ¥å¼îƒ¬îŠ¦å¹•îŒ¥åˆå…€/p> The cloud reflex arcs of the Internet cloud have actually appeared around us today. Almost every moment, the phenomenon of Internet reflexes initiated from all over the world is continuously generated and disappeared. For example, if a car sensor finds a thieves, it sends a text message to the owner, and the owner rushes to grab the thieves. The humidity sensor finds that the air has increased in humidity and signs of rain, and informs the field excavation equipment to turn on the rainproof equipment. Corresponding to the human body's nerve reflex arc, we can see from the foregoing: 1. The refractors of the cloud reflex arc consist of networked sensors (including cameras) 2. The effectors of cloud reflex arc are mainly composed of networked office equipment, smart manufacturing, smart driving, smart medical and so on. 3. The central nervous system of cloud reflex arc is the Internet's central nervous system (cloud computing + big data + artificial intelligence). Edge calculations will enhance the intelligence and reaction speed of the cloud's reflex arc sensors and effectors. As a result of the combination of the Internet system and artificial intelligence, the cloud reflex arc will play a very important role in the future development of the Internet. From a practical point of view, there are a total of 9 different types of cloud reflex arcs. The maturity of these reflex arcs depends on the further integration of the Internet and artificial intelligence technologies. The first is a cloud-reflection arc from the sensor to the smart device (A->F in the figure): For example, in a building, when the temperature sensor detects that the room temperature has risen above a certain temperature, the gas sensor detects that the indoor carbon dioxide concentration rises. High, then the alarm information is transmitted to the server center through the Internet. The server sends instructions to the building fire extinguishing robot. The floor fire extinguishing robot controls the water gun to extinguish the fire. The second is a sensor-to-person cloud reflex arc (A->B in the figure): For example, in the building, the temperature sensor detects that the room temperature has risen by more than 100 degrees, and the gas sensor detects that the indoor carbon dioxide concentration has increased. Then the alarm information is transmitted to the server center through the Internet line, the server sends the information to the nearby fire brigade, and the fire brigade dispatches firefighters to the large building to implement the fire extinguishing. The third is the sensor-to-intelligent program's cloud reflex arc (A->D in the figure): For example, in the building, the temperature sensor detects that the room temperature has risen by more than 100 degrees, and the gas sensor has detected an increase in the indoor carbon dioxide concentration. Therefore, the alarm information is transmitted to the server center via the Internet line, and the server sends information to the AI ​​neurons in the Internet neural network, which is an intelligent program in a large social network, and the intelligent program determines whether or not the danger level is reported. The fourth is the cloud-reflection arc of intelligent programs to smart devices (cloud reflection architecture diagram D->F): For example, an automatic monitoring program running on an Internet server detects changes in the capacity of the server data space of a cloud computer room in a suburban city. When the program finds that the data space is full, it sends an alarm message to the Internet center server. The central server issues an instruction to start the standby machine in the cloud computer room and expands the data space. The fifth type is a cloud-reflection arc of an intelligent program to a person (C->E in the figure): For example, an automatic monitoring program running in an Internet server detects changes in the capacity of a server data space in a suburban cloud computer room, when the program finds a data space. When it is full, an alarm message is sent to the server of the Internet center. The central server issues a short message or an e-mail to remind the on-duty personnel of the engine room to start the backup machine in the cloud computer room and expand the data space. The sixth type is a cloud-reflection arc from an intelligent program to an intelligent program (C->D in the figure): This type of neural reflex arc can be viewed as a dialog of the cloud-based artificial intelligence system. For example, an automatic monitoring program running on an Internet server detects a change in the capacity of a server data space in a suburban cloud computer room. When the program finds that the data space is full, an alarm message is sent to an Internet center server, and the central server issues an instruction to the cloud computer room. The maintenance program stops writing data to the data space and avoids data space overload. The seventh type is a cloud-reflection arc (B->F in the figure) from a person to a smart device: For example, in a building, the on-duty staff of the building monitoring room found that an office had flames and smoke, and then pressed the alarm button to send an alarm message. It is sent to the server center through the Internet. The server sends instructions to the building fire extinguishing robot. The floor fire control robot controls the water gun to extinguish the fire. The eighth is a human-to-person cloud reflex arc (cloud reflection architecture diagram B->E): For example, in the building, the building supervisory staff on duty in the computer room found an office fire and smoke, then press the alarm button to alarm The information is transmitted to the server center through the Internet. The server sends information to the nearby fire brigade. The fire brigade dispatches firefighters to the large building to implement the fire extinguishing. The ninth type is the cloud-reflection arc of a human-to-intelligent program (cloud reflection architecture diagram B->D): For example, in the building, the on-duty personnel of the building's monitoring and control room discovered an office fire and smoke, and pressed the alarm button. The alarm information is transmitted to the server center via the Internet. The server sends information to the AI ​​neurons in the Internet neural network, which is an intelligent program in a large social network. The intelligent program determines whether the risk level is reported or not. IV. Application Cases and Future Impact of Cloud Reflex Arc In practical applications, the cloud reflex arc of the cloud has gradually appeared in front of people. For example, the fire department in Wuxi started to use the remote monitoring and rescue system for home fires to help the left-behind elderly and left-behind children in rural areas to carry out fire warning and rapid response. This system consists of wireless terminals, service platforms, and sensing devices (smoke, emergency button, etc.). Its working process is a typical cloud reflex arc. When a fire or other emergency occurs, the detector sends an alarm signal (or manually presses the rescue button). The fire alarm information will be transmitted to four places through the GPRS or TD wireless network: First, the community (cell) monitoring center; the second is the area The fire patrol car in the fire department; the third is the 119 fire dispatch command center in the city; and the fourth is the mobile phone of the family and their loved ones and neighbors' friends on the fire. When the four parties received the alarm, they could rush to the scene for assistance at the first time, winning time for extinguishing the fire and emergency disposal, avoiding the spread of fire, and minimizing casualties and property losses. From this system of Wuxi fire fighting we can see that in addition to sensors and mechanical devices that can act as sensors and effectors, people are also important factors in the Internet's reflex arc, either acting as sensors or effectors. The development of cloud reflex arc is an inevitable outcome of the combination of the Internet + AI, and its development will have a wide range of impacts on the Internet-based artificial intelligence technology, the new business model of the Internet, and the construction of smart cities. Bibliography: "The Evolution of the Internet as Viewed from the Structural Mechanism of the Human Brain", Liu Feng, Peng Geng, Liu Ying, Ergonomics, 2008 Pay attention to NetEase smart public number (smartman163), get the latest report of artificial intelligence industry.
