Feb. HKUST

1. The first novel coronavirus pneumonia potential drug was approved to be put on the market.

Hot wire!
The first potential novel coronavirus pneumonia was approved by the State Food and drug administration.
Previously, the Ministry of science and technology officially announced three potential Western anti epidemic drugs: fabiravir, chloroquine phosphate and ridcivir.
Both of them showed obvious curative effect and low side effect in clinical trials.
Now, fapiravir is first approved by Zhejiang Haizheng pharmaceutical industry to start production. Novel coronavirus pneumonia is a new indication of influenza, but the flu is approved after the launch.
The novel coronavirus trial for novel coronavirus pneumonia is continuing.
In addition, in addition to the new progress of therapeutic drugs, the good news also comes from the detection aspect. Under the guidance of academician Zhong Nanshan, a more rapid antibody detection kit was developed to shorten the detection results to 15 minutes.
The battle against the epidemic may be dawning ahead.
What is fapiravir?
Fabiravir, one of the three drugs focused by the research team.
Also known as favipiravir or avigan in English, developed by Toyama chemical company of Japan.
This is an experimental anti influenza drug, which belongs to a broad-spectrum anti RNA virus drug, and has activity for many RNA viruses. In 2014, it was approved in Japan as a storage drug for use in anti influenza pandemics.
Previous trials have shown that fabiravir has a certain effect on Ebola virus. In 2015, the results showed that it can reduce the mortality of patients with low to moderate Ebola virus levels in the blood.
Fabiravir can also effectively inhibit Ebola virus, yellow fever virus, chikungunya virus and norovirus. A new study shows that the EC50 of the new coronavirus in vitro cell line experiment is 61.88 μ M.
In the subsequent novel coronavirus pneumonia trial, the clinical efficacy of the drug was gradually demonstrated.
The conclusion disclosed by Shenzhen three hospitals involved in the clinical practice is that: fabiravir is safe and effective, its antiviral effect is better than that of klitschke, and its side effects are also significantly lower than that of klitschke. It is suggested to expand the scale of clinical application.





On February 15, at the press conference of the joint defense and control mechanism of the State Council, Zhang Xinmin, director of the biological center of the Ministry of science and technology, further disclosed the clinical results:
Fabiravir is an overseas drug for the treatment of influenza. At present, clinical trials have been carried out in Shenzhen, with more than 70 patients in the group, including the control group, initially showing a relatively obvious efficacy and low adverse reactions. On the 3rd to 4th day after treatment, the negative rate of viral nucleic acid in the treatment group was significantly higher than that in the control group.
Novel coronavirus pneumonia was approved by the State Food and Drug Administration on the same day. It became the first drug approved for sale in the country during the epidemic period.
Haizheng pharmaceutical issued a notice on the morning of February 17, saying that it had obtained both the drug registration approval and the drug clinical trial approval.
According to the approval document for drug registration, the approved fabiravir is a tablet produced by Zhejiang Haizheng Pharmaceutical Co., Ltd. and is mainly used for the treatment of new or re epidemic influenza in adults.
Although the approved drug is influenza, clinical trials of the drug against the new coronavirus have begun.
The approval results of clinical trial approval document show that:
The novel coronavirus pneumonia is adapted for clinical trial.
Previously, Haizheng pharmaceutical had said in an interview with the medical media, the magic cube of Medicine:
Once approved, the novel coronavirus pneumonia can be used by clinical experts to treat pneumonia in a small area. If more novel coronavirus pneumonia is effective in treating new crown pneumonia, more drugs will be provided through appropriate channels to benefit patients.
Novel coronavirus pneumonia trials and approval of clinical trials are also the reflection of the accelerated drug treatment review of potential treatment of new crown pneumonia. It took only half a month for the drug to be put on the market and approved for production. The clinical trials were promoted simultaneously.
We look forward to further good news from the clinical trials of fabiravir.
Progress of other two drugs in clinical trials
One is known as readcivir ("people's hope"), the other is chloroquine phosphate.
Novel coronavirus pneumonia research has been put into clinical trials as of February 16th, according to Zhang Xinmin, director of the Biotechnology Center of the Ministry of science and technology. As of February 16th, there were 136 clinical trials of new coronavirus pneumonia.
But the main focus of the research team is on chloroquine phosphate and rendisivir, as well as fabiravir, which is now in production.
What are the first two drugs? We started to know each other.
First, chloroquine phosphate, an antimalarial drug, has been on the market for many years.
Its antiviral mechanism is mainly as follows:
1. Because chloroquine can change the pH value of endocytosis, it has a significant inhibitory effect on virus infection that invades cells through endocytosis, such as Borna disease virus, avian white blood virus, Zika virus, etc.
2. Chloroquine can affect virus replication by inhibiting gene expression. In vitro and in vivo experiments showed that chloroquine could change glycosylation mode of gp120 envelope and inhibit HIV replication in CD4 + T cells.
3. Chloroquine, as a good autophagy inhibitor, can interfere with virus infection and replication by influencing autophagy.




The animal experiments showed that chloroquine could effectively inhibit autophagy in the lungs of H5N1 mice and reduce the damage of alveolar epithelium. It has also been reported that chloroquine can block the autophagy induced by Zika virus and thus inhibit the replication of the virus. In mice, it has also been shown that chloroquine can cut off the vertical infection of Zika virus from mother to fetus.
In addition, in 2004, it was reported that chloroquine could inhibit the replication of SARS virus in vitro.
Zhang Xinming said that this time, in vitro experiments showed that chloroquine phosphate had a good inhibitory effect on the new coronavirus. And because it is a listed drug, its safety is also guaranteed.
At present, clinical research is being carried out in more than ten hospitals in Beijing and Guangdong, with more than 100 patients enrolled. Recently, the multicenter clinical trial of chloroquine phosphate will be launched in Hunan Province.
And then there's redceway, which more people know.
Radcivir, a nucleotide analogue prodrug, can inhibit RNA dependent RNA synthetase (RdRp), which was developed by Gilead science company, the world's first antiviral drug research company, and has been developed for nearly 10 years.
In vitro and animal models, radcivir is effective against SARS and mers, so experts judge that radcivir is also effective against new coronavirus.
In the first confirmed novel coronavirus pneumonia patient in the United States, Reed's Wei Wei was used in the treatment. Since then, redceway has been given further hope.
At present, radcivir is still in the clinical trial stage. Wuhan Jinyintan hospital, Wuhan Pulmonary Hospital, Tongji Hospital, Xiehe Hospital, etc. have carried out drug trials.
168 severe patients and 17 light and common patients have been enrolled.
△ Academician Wang Chen (first from the right), vice president of Chinese Academy of engineering, introduced the project. Picture from Xinhua News Agency
According to the latest news of Changjiang Daily, from February 5, a 68 year old severe male patient in Wuhan Jinyintan hospital has become the first probationer of radcivir, and is now progressing smoothly.
Professor Zhong Ming, deputy director of the Department of critical medicine of Zhongshan Hospital Affiliated to Fudan University, also said in a video interview:
Including the doctors at the clinical front line, they still have great hopes for this drug. But there is no data to tell us whether the drug works or not. The drug takes 10 days and is double-blind. The first group of patients who used it could only see the real results after 10 days of Unblinding.
Professor Zhong Ming is the first group of medical staff in Shanghai to assist Hubei Province. At present, he is treating severe patients in Wuhan Jinyintan hospital.


Peng Zhiyong, director of the intensive care unit of Central South Hospital, said that the final effect of radcivir should be evaluated by experts after the end of the trial. There are many indicators, including virological evidence, clinical indicators and imaging indicators for comprehensive judgment.
The clinical trial was originally planned to end on April 27.
Test kit directed by Zhong Nanshan: 1 drop of blood 15 minutes results
In addition to drugs, there are also new developments in detection kits.
According to CCTV, one drop of blood can be collected from the rapid test kit developed under the guidance of Zhong Nanshan and the test results can be obtained within 15 minutes.
Novel coronavirus IgM antibody rapid test kit was developed by China National Academy of Sciences, Guangzhou Institute of biomedicine and health, Guangzhou regenerative medicine and Guangdong provincial laboratory in February 14th, and has been evaluated in laboratory and clinic by the National Key Laboratory of respiratory diseases, academician Zhong Nanshan.
Novel coronavirus IgM antibody was detected by indirect immunoassay with colloidal gold immunochromatography. It is expected that the detection results will be obtained within 15 minutes by taking a single drop of blood, and the positive bands can be detected after the plasma is diluted 500 to 1000 times.
A novel coronavirus infection test (PCR negative test) was performed in a hospital in Hubei. The kit was able to detect positive (IgM) positive results, which could be complementary to nucleic acid detection.
At present, the novel coronavirus has been sent to the grass-roots health institutions such as Wuhan, Huanggang, Daye, Hubei, etc., and is combined with nucleic acid detection technology to test and assess the infection of new coronavirus.
The researchers said that this method is expected to break through the limitations of the existing detection methods on people and places, shorten the detection time, realize the rapid diagnosis of suspected patients and the on-site screening of close contact population, and promote the forward and downward movement of diagnostic screening.
Therefore, with the acceleration of testing and the good news from drugs, the day of victory in the fight against the epidemic may also continue to accelerate.
Among all things, hope is the most beautiful.
In addition to praising the front-line medical angel, the front-line scientific research personnel who are quick to report, are also worthy of admiration.
One more thing
Finally, after the first production of fabilavir, many netizens also paid attention to the progress of traditional medicine in anti epidemic.
At present, on the website of China clinical trial registration center, there are many traditional Chinese medicine and proprietary Chinese medicine (including Shuanghuanglian oral liquid, atomized inhalation Xiyanping injection, etc.) are also carrying out clinical trials for new coronavirus. The specific data and results have not been disclosed, and the specific efficacy is unknown.
However, some netizens also provided the scientific basis for traditional Chinese medicine to be effective against new coronavirus.
In a recent review paper, NEJM (New England Journal of Medicine), an international authoritative medical journal, focused on placebo effect and anti placebo effect.



The paper points out that placebo effect and anti placebo effect are the effects of patients' positive and negative expectations of their own health. Placebo effect leads to favorable outcome, while anti placebo effect leads to harmful and dangerous outcome.
In clinical practice, placebo and anti placebo effects are powerful, common and common. These effects can be produced by neurobiological mechanisms, treatment-related information, patient expectations, previous experience with drugs or procedures, and the treatment environment.
In clinical practice and clinical trials, strategies to promote placebo effect and avoid anti placebo effect can improve treatment outcome and reduce unintentional symptom deterioration.
A more straightforward interpretation is that in front of the virus that attacks the immune system, having a placebo can make patients more confident to stimulate the power of the autoimmune system, and ultimately play the role of anti-virus.
Of course, for more details, it is advisable to move the transmission door:
Address of Chinese paper:
https://nejmqianyan.cn/article/YXQYra1907805
In addition, there are two new progress of coronavirus research.
First, the new coronavirus may not only damage the respiratory system, such as the lungs, but also the kidneys and testicles. Therefore, people who eat wild game pay attention to that wild game not only has no scientific basis, but also may backfire.
Second, a study from the University of South Carolina suggests that smokers may be more susceptible to the new coronavirus.
For details, please refer to other pictures and texts pushed today.
I wish you all good health, continue to do a good job of protection, victory in the epidemic, is not far away~~





2. AI doctor "war epidemic" is on the front line



In case of emergency rescue, AI doctor also sent out.
 
As of February 6, with the continuous development of new coronavirus pneumonia, there have been 31161 confirmed cases and 26359 suspected cases in China. However, due to the high shortage of medical resources, especially the structural lack of rapid diagnosis and treatment capacity in the core epidemic areas.
 
Novel coronavirus pneumonia can be effectively screened and diagnosed by AI technology, which helps to identify potential patients and strengthen prevention and control.
 
Before that, Alibaba, Baidu and other technology giants took the lead in the battle of AI technology epidemic prevention. Aiming at the novel coronavirus AI diagnostic technology, the AI hospital developed the algorithm, which can shorten the initial hours of suspected case gene analysis to half an hour, significantly shorten the time of diagnosis, and accurately detect the variation of the virus. Baidu has opened the novel coronavirus RNA spatial structure prediction speed freely, opening the linear time algorithm LinearFold and RNA structure prediction website.
 
At the same time, some AI innovation and entrepreneurship companies can not be ignored.
 
Novel coronavirus infection was launched in February 4th by the Ministry of industry and commerce, which is an initiative to the AI Association, alliance, enterprises and institutions. But in fact, many AI companies aiming at Ai + medical market have provided AI solutions for assistant doctors based on the needs of first-line hospitals earlier.
 
Since the outbreak of the epidemic, the R & D departments of these AI + medical field companies have worked overtime for several consecutive nights, launched corresponding solutions in a short time, and maintained rapid iteration, and launched emergency online in front-line medical institutions.
 
Novel coronavirus pneumonia has also become a great opportunity to test the application of AI technology as a potential application scenario for AI technology.
 
Open technology
Fight for ten days continuously, launch AI temperature measurement system
 
Novel coronavirus pneumonia is a major indicator of whether a person is infected with a new crown pneumonia. In public places with dense flow of people, the key is how to achieve "non-contact induction", "high-efficiency passage" and "high temperature intelligent early warning".
 
After the outbreak, Kuangshi technology mobilized nearly 100 R & D teams in response to relevant needs to conduct in-depth discussion and communication on the requirements of fever screening system, solution function planning, interface scheme, test joint commissioning, supply chain integration and other issues. Finally, after ten days of hard work, based on the needs of landing scene, Kuang Shi R & D team put forward an innovative solution of "human body recognition + human image recognition + infrared / visible light dual sensor", identify the high-temperature personnel in the flow of people through the front-end infrared camera, and then according to the human body and face information of the suspected fever, through AI Technology can help workers in various public places and other high-density personnel flow scenarios to quickly screen people with abnormal temperature, and solve the traffic efficiency and controllability in open places after the epidemic is stable.   


 
According to Kuang Shi, compared with the artificial short-range detection method and the traditional fever screening system, Kuang Shi will focus on the research and development of the temperature measurement accuracy of long distance and the availability of high-density human flow.
 
First of all, Kuang Shi carried out a special model optimization for masks. Even when masks and hats cover a large area of the face, the system can help the staff quickly screen the passing crowd, with the recognition error of ± 0.3 ℃. The public does not need to take off the protection or queue to gather, greatly improving the safety and detection efficiency of the public space.
 
Another major feature of the system is that it can support the non-contact remote temperature measurement of more than 3 meters. Once there is a suspected heating personnel, it will automatically alarm. Combined with the human Reid detection and retrieval technology developed by Kuangshi, the system can help the staff to quickly screen the location clues of fever personnel, so as to further confirm the detection and medical observation.
 
Kuangshi R & D personnel said that the system's intelligent suspected high heat alarm bandwidth can reach 15 people per second, and a set of system can deploy 16 channels, basically ensuring a subway port control. In this way, with the help of AI system, only one staff member can control the site for the channel with large traffic, which greatly reduces the risk of infection for the front-line staff.
 
Shang Tang Technology
Novel coronavirus pneumonia screening, diagnosis and treatment of three major AI programmes
 
Novel coronavirus pneumonia is the key point in the epidemic prevention work in public places.
 
For this reason, Shangtang technology, using face recognition algorithm and thermal imaging intelligent temperature measurement technology, has rapidly launched the "Ai intelligent epidemic prevention solution" in the shortest time, which mainly includes area access module and entrance access module, which can realize the recognition and management of personnel temperature, mask wearing, personnel identity and other aspects, and can display and archive real-time information in a unified way.
       
Meanwhile, novel coronavirus pneumonia has been upgraded rapidly to the rescue of many new hospitals and cooperative medical institutions based on the existing SenseCare intelligent diagnosis and treatment platform.
       
Through AI, we can automatically detect, locate and analyze the lesions of suspected patients' lung CT images, and generate automatic reports, so as to help medical staff quickly screen suspected new crown cases, isolate and further diagnose them, so as to reduce the risk of cross infection. At the same time, the follow-up function of the product can also help doctors to accurately evaluate the treatment effect of patients with AI To provide the front-line medical workers with intelligent assistance that can be traced and supported.
 
In addition, in order to better cope with the future variation of viruses and make up for the problems that small molecule drugs are prone to drug resistance, the smart health team of Shangtang, based on the long-term research accumulation in AI pharmacology and AI gene analysis, has carried out relevant research on the "new crown" biological macromolecule treatment scheme in combination with many professional medical institutions and scientific research universities, hoping to provide more To solve the problem of drug resistance of virus to single drug and improve the resistance to virus mutation, multiple treatment methods and treatment methods are needed to fight against virus and prevent further spread of infected population.




According to technology
The novel coronavirus pneumonia intelligent image evaluation system completes quantitative analysis within 3 seconds.
 
According to the novel coronavirus pneumonia diagnosis and treatment plan (trial version fourth), the diagnostic criteria of suspected cases, pneumonia imaging features are one of the important diagnostic and disease assessment indicators.
 
Perhaps novel coronavirus pneumonia evaluation system is released directly by its official website, which is released by its public official clinical center in Shanghai, compared with other companies' press releases.
       
The novel coronavirus novel coronavirus pneumonia image evaluation system, which is based on medical development, is realized by image algorithm to realize the intelligent diagnosis and quantitative evaluation of the new coronavirus pneumonia CT image, and to classify the severity of various types of lesions, diffuse lesions and all lung involvement, and through the key image characteristics of the lesion's shape, range and density. Quantitative and histochemical analysis can accurately calculate the accumulated pneumonia load of the disease, and realize the dynamic 4D comparison of the whole lung disease of CT, which is helpful for clinical judgment of the condition, evaluation of the curative effect and prediction of the prognosis.
 
For novel coronavirus pneumonia, the classification and curative effect evaluation of pneumonia severity is the focus of clinical attention. The method of quantitative evaluation of lesions involves many factors such as accumulated lung volume, density, etc. at present, there is no unified standard. In the past, traditional manual ROI method is often used for quantitative evaluation, which takes 5-6 hours, with low efficiency and difficult clinical promotion. However, the AI system according to the figure can realize the automatic detection of lesions, and the quantification can be completed within 2-3 seconds Analysis, greatly improving the efficiency of accurate quantitative analysis.
 
Comparing novel coronavirus pneumonia with AI system, the Shanghai Public Health Clinical Center pointed out that the quantitative analysis of the system is similar to the result of the doctor's evaluation (R=0.87 (p<0.001) in correlation study), showing the high stability of the diagnostic quality.
       
       
In addition, novel coronavirus pneumonia is also available on the basis of "knowledge of new coronary pneumonia", which is convenient for people to get knowledge of the epidemic prevention and control and help identify the suspected cases in time.
Imagination Technology
First AI special edition of pneumonia with three abilities
The technology of novel coronavirus pneumonia has also been introduced by AI.
 
They put forward three abilities of pneumonia AI Special Edition: imaging doctor + AI can quickly respond in special period and give diagnosis report, screen and screen "highly suspected pneumonia patients", reduce patients' queuing time and risk of cross infection in Hospital; through CT + AI, it can help speed up the identification of infected persons in primary medical institutions, so as to quickly isolate, diagnose and treat them CT + AI has better operability, which can quickly and directly identify asymptomatic patients and early patients. It is an effective guarantee for fundamentally eliminating the epidemic situation.
 
The intelligent processing of AI can be inferred, such as detection of pulmonary infection focus, analysis of abnormal cases and severe cases, orderly and graded treatment, coordination of medical resources, full-automatic comparison of front and back films and evaluation of curative effect, etc., which can help doctors to complete treatment evaluation more quickly and accurately, and save energy and time for doctors who are overloaded.
       
 
It is assumed that AI special edition of pneumonia has been put into use all over the country, such as Wuhan, the central city of the epidemic. With the in-depth use of AI in clinical practice and the screening tool of CT + AI assisted diagnosis, some findings of AI in clinical diagnosis and treatment are proposed through practical cases
 
*CT prediction was earlier than RT-PCR, and AI prediction was consistent with the final diagnosis;
*Compared with the front and back films, the lesion area was not significantly changed, but the lesion density was significantly increased compared with the front film, indicating the progress of the lesion. It was assumed that the AI prediction results were consistent with the clinical symptoms, and the patient was finally transferred to hospital for isolation treatment;
*The special edition of AI pneumonia is supposed to provide accurate quantitative analysis tools for clinical practice and accurately judge the course of disease.



Shen Rui medical treatment
Novel coronavirus pneumonia enhanced version AI system, output structured report
Since novel coronavirus pneumonia is needed for screening, Shen Rui medical has received two sets of Dr.Wise AI lung medical diagnostic assistant system and Shanghai CT demand for cooperation with Hubei.
 
Based on novel coronavirus pneumonia novel coronavirus infection, novel coronavirus pneumonia is diagnosed by Dr.Wise AI, and the Dr.Wise intelligent lung disease intelligent solution (new crown pneumonia Enhanced Edition) is used to detect various signs of different types of pneumonia. Meanwhile, the detection of similar signs of new crown pneumonia is strengthened, especially for localized or multiple cases. Under the pleura and in the middle and outer zone of bilateral lung fields, there are high detection rates of spot shadow, ground glass shadow, grid shadow and strip shadow.
 
The novel coronavirus report is also classified into various categories, including intelligent classification of various signs of pneumonia, quantitative analysis of consolidation and ground glass opacity, auxiliary doctors to determine the stage and severity of pneumonia, and providing early warning clues for suspected pneumonia.
    
In addition, the novel coronavirus pneumonia solution (new enhanced pneumonia Enhanced Edition) provides the whole process of diagnosis and treatment management, including location, quantitative analysis, multi time follow-up of CT lung inflammation, standard structured report, etc., which helps doctors improve the diagnostic efficiency greatly and improve the accuracy of Dr.Wise diagnosis. Whether the suspected cases are confirmed as imaging diagnosis cases, or the patients in the transition period are confirmed, multiple CT Reexamination and follow-up are essential.
The "intelligent follow-up system for lung diseases" can carry out intelligent analysis for the follow-up examination of the same patient, compare the changes of lesions, and assist doctors to accurately judge the progress of the disease.
Joint shadow intelligence
Intelligent Tianyan CT support, reduce doctor-patient contact
In traditional scanning, radiologists need to contact patients closely in body position selection, patient positioning and scanning range determination. In this special period, how to reduce the close contact between doctors and suspected patients to complete the scan?
 
The "intelligent eye" of UAI vision is designed to be carried on medical imaging equipment to complete full dynamic capture of 360 ° human body in real time, generate 3D human image automatically, and realize "one button intelligent" scanning. UAI vision can intelligently identify the patient's age, height, body shape and other information, and accurately push the parameters required for scanning to the doctor. Automatic recognition of patient positioning, real-time remote "remote control" scanning equipment, automatic move to the position to be photographed, realize compartment operation, maximize isolation of doctors and patients, greatly reduce the scanning workload.
       
 
The "intelligent eye CT" combined with SkyEye can realize intelligent positioning and accurate recognition. Its automatic monitoring system can confirm the scanning range of the patient's chest through the face detection function, realize automatic positioning, and reduce the contact between doctors and patients. At the same time, after the face detection is initially located in the chest area, SkyEye CT can quickly scan the ultra-low dose CT location image (topogram), and automatically extract the position of the lung in the location image through segmentation technology, so as to accurately determine the scanning range, so as to reduce the unnecessary radiation in the scanning process.
       
Image segmentation of novel coronavirus pneumonia chest CT image (left) and pneumonia infection area (right)
 
Novel coronavirus is being co operated with some epidemic prevention hospitals, using uAI's original segmentation algorithm to assist in the diagnosis and quantification of new coronavirus pneumonia. The novel coronavirus novel coronavirus region can be detected from the CT image by the technology. The area is accurately segmented and its volume and composition are quantified and analyzed. The report provides accurate quantitative basis for early diagnosis of pneumonia, and helps doctors quickly identify and diagnose new coronavirus infection.
 
Summary
 
In general, AI plays an important role in energy saving of primary screen ring in epidemic prevention. Due to time constraints, major companies are busy with product iterations. At present, no detailed data has been released on the specific accuracy indicators of the "Ai doctor" system.
 
Although there are challenges, this "war epidemic" is also an opportunity for the development of AI + medical industry.
 
Like human doctors, they will accumulate corresponding medical experience after the epidemic is eliminated. We also hope that the technical solutions AI company has invested in this epidemic prevention will accumulate experience in the future. Novel coronavirus pneumonia is not only a response to the larger dimension, but also hopes that AI will play a more significant role in other medical applications and promote the development of human society in the future.




3. Human machine integration Intelligence: artificial intelligence 3.0

Absrtact: artificial intelligence (objective Intelligence) and human intelligence (subjective Intelligence) are most likely to evolve into a pair of opposites of yin and Yang, which are mutually matched, thus opening a new era, that is, the era of man-machine integration intelligence. In the era of human-computer integration intelligence, the role of organization and management may be more critical, because the object of management has been upgraded to a machine with extremely powerful functions and certain human intelligence, as well as organizations and individuals equipped with powerful capabilities by intelligent machines, and the management environment has thus been transformed into a complex giant system formed by intelligent human-computer interaction.
(Li Ping: Li Dasan, chief Professor, University of Nottingham, Ningbo; Yang Zhengyin, research director, Dahua District, American Innovation Leadership Center; Ph.D. student, University of Nottingham, Ningbo)
Human beings have created their opponents in the objective field, namely artificial intelligence (AI) in the field of subjective intelligence, which is considered to be their own unique characteristics. No matter how artificial intelligence evolves in the future, human beings will undergo revolutionary changes due to the development of artificial intelligence, and the structure of social organizations in the future will also become a new scene. We boldly predict that artificial intelligence (objective Intelligence) and human intelligence (subjective Intelligence) are most likely to evolve into a pair of opposites of yin and Yang, i.e. "man-machine integration intelligence". This will open a new era, that is, "the era of human-computer integration intelligence". To this end, we should start from now on to make the necessary comprehensive preparations, especially in terms of organizational structure.
Current situation and controversy of AI development
As a concept, or a new discipline, the history of artificial intelligence is only 62 years, not very long. When John McCarthy proposed the concept of "artificial intelligence" at the Dartmouth conference in 1956, he probably didn't expect that the word would attract so many discussions and disputes in the next half century. When it provided people with unlimited imagination space for animation, science fiction and so on, it also brought many people's deep doubts. With the great progress of human science and technology in the past decades represented by information technology, artificial intelligence has also experienced at least two upsurges.  
The first wave, artificial intelligence 1.0, occurred between 1956 and 1974. Scientists have introduced symbolic methods into statistical methods for semantic processing, and knowledge-based methods have emerged. At the same time, scientists have invented a variety of algorithms with great influence, such as the embryonic Bellman formula of deep learning model. In addition to new advances in algorithms and methodology, scientists have also made machines with preliminary intelligence. Artificial intelligence 2.0 is from the early 1980s to the end of the 20th century. A series of significant achievements have been made in the mathematical models related to artificial intelligence, such as the famous multilayer neural network and BP back propagation algorithm, which further gave birth to a highly intelligent machine capable of playing chess with human beings. Other achievements include the machine that can automatically identify the post code on the envelope through the artificial intelligence network, with the accuracy of more than 99%, which has exceeded the level of ordinary people. At the same time, Carnegie Mellon University has developed an expert system for DEC company, which can help DEC company save about 40 million dollars every year, especially in decision-making. Many countries, including Japan and the United States, once again invested heavily in the development of so-called fifth generation computers, which were called artificial intelligence computers at that time. In AI 1.0 and 2.0, due to the computer technology in hardware and software (especially the former) failed to achieve breakthrough progress, the relevant applications are always difficult to achieve the expected results, unable to support large-scale commercial applications, and finally after two high tides and low valleys, AI fell silent.

At present, the third wave, artificial intelligence 3.0, has been set off since the beginning of the 21st century. With the rapid development of information technology and the rapid popularization of the Internet, marked by the 2006 deep learning model, artificial intelligence ushered in the third high-speed growth and entered a new stage of the development of artificial intelligence. With the popularity of the Internet, the ubiquity of sensors, the emergence of big data, the development of e-commerce and the rise of information communities, data and knowledge are converging and interacting in human society, physical space and information space, and the information environment and data base in which the development of artificial intelligence is located have undergone tremendous and profound changes, which constitute the driving force for the new trend of artificial intelligence The external driving force of the stage. At the same time, the goals and concepts of artificial intelligence have been adjusted significantly. New breakthroughs in scientific basis and implementation carrier indicate the growth of internal power, including the huge progress of computer in hardware and software technology (especially the former), such as brain like computing, deep learning, reinforcement learning and a series of technologies. Therefore, artificial intelligence has entered the 3.0 stage of its development. As AI is still in the early stage of development, the definition of AI is still divergent, there is no unified authority, and it is still in the process of improvement. According to the definition of artificial intelligence in the new generation of artificial intelligence development white paper (2017) compiled by China Electronics Society, it refers to the evolution from computer simulation of human intelligence to helping guide and improve human intelligence, through promoting the interconnection and integration of machines, people and networks, more closely integrating into human production and life, and evolving from auxiliary equipment and tools to collaborative and interactive assistants and partners Companion. It is also believed that artificial intelligence should refer to the integration of computer programming, neural network and problem solving. In 1950, Alan Turing, a British scientist, put forward "Turing test", that is, whether a robot has intelligence or not depends on whether people can distinguish it from a robot after communicating with it. Therefore, the standard of "intelligence" started around a robot with independent thinking ability.  
With different levels of humanoid intelligence, it can realize the closed-loop workflow of "perception decision behavior feedback", assist human production, serve human life, and automatically perform various kinds of machine devices, mainly including intelligent industrial robots, intelligent service robots and intelligent special robots. According to the current general classification of artificial intelligence, artificial intelligence may evolve into three levels: (1) primary artificial intelligence refers to the robot with partial intelligence of the owner; (2) advanced artificial intelligence refers to the robot with all intelligence except human's self-consciousness, also known as strong artificial intelligence or general artificial intelligence; (3) super artificial intelligence refers to the soul given to human beings ——Including emotional, ethical and moral factors - living robot, also known as human artificial intelligence. At present, the research and application of artificial intelligence are mainly focused on low-level artificial intelligence (such as single or special intelligence of alpha dog type, but lack of universal intelligence possessed by human beings); some scholars have begun to explore high-level artificial intelligence (such as cross domain universal intelligence possessed by human beings), but there is still a long way to go from the practical application, and there is no such thing There is little debate (for example, whether the development direction of advanced artificial intelligence is to replace human intelligence, or to form integrated intelligence through human-computer cooperation); as for super artificial intelligence, because of the deep-seated issues involving philosophy and ethics, the parties to the debate may not reach a consensus for a long time.  
Before people can imagine the new era of artificial intelligence 3.0, they suddenly find that artificial intelligence has a close relationship with human life. For example, because the new generation of artificial intelligence accelerates the replacement of machines and people, leading to the increasingly severe trend of structural unemployment. In addition, privacy protection has become a difficulty, and the definition of data ownership, privacy and permission has become a challenge War. The most important issue is that people have different understanding and expectation on the future development direction of artificial intelligence, especially on whether artificial intelligence can replace human intelligence and whether it threatens human survival. These differences in people's understanding of artificial intelligence will also have a key intervention in the development of artificial intelligence and its potential impact on the future society.

Human computer interaction and cooperation: the key of AI 3.0
We advocate that the main direction of artificial intelligence should be man-machine cooperation, rather than simply replacing human beings with machines. Therefore, the fusion dual intelligence produced by human-computer interaction can become the top priority of the future development of artificial intelligence. It aims to improve the performance of artificial intelligence system through human-computer interaction and collaboration, make artificial intelligence become the natural extension and expansion of human intelligence, and solve complex problems more efficiently through human-computer collaboration, which has profound scientific significance and huge industrialization prospect.
If the previous development of artificial intelligence focused on data input and processing, for example, the three main signs of the development history of artificial intelligence: dark blue, Watson and alphago, which are all related to data and are dealing with a large number of data, rules and planning in the past, then the next important direction of the development of artificial intelligence in the future will be the intelligent system of human-computer interaction. Because the three systems mentioned above all have a bottleneck problem, that is, they can only "get shape and forget meaning", not "get content and forget shape". People's intelligence needs to make decisions in time, instead of computing according to given rules like computers or current weak artificial intelligence. Facing the complex reality, people need to innovate "algorithm" on the spot. Therefore, artificial intelligence will gradually approach to human intelligence, imitating human beings' comprehensive use of vision, language, hearing and other perceptual information, to achieve recognition, reasoning, design, creation, prediction and other functions. In the human-computer interaction intelligent system, the automatic recommendation, autonomous learning, natural evolution and autoimmunity are the most important four aspects. According to the latest development trend, the new generation of artificial intelligence will have these four elements: large amount of data, computing power, algorithm model and multiple application scenarios.  
This kind of transformation has become the focus of human-computer interaction from simply simulating human intelligence with computer, building a single agent with perceptual intelligence and cognitive intelligence, to building a multi-agent collaborative group intelligence. Group intelligence fully embodies the idea of "overall consideration and overall optimization", which has the advantages of decentralization, self-healing and efficient information sharing. Related group intelligence technology has begun to sprout and become a research hotspot.  
Human computer cooperation is giving birth to a new form of fusion intelligence. Human intelligence has unparalleled advantages in perception, reasoning, induction and learning, while machine intelligence is ahead of human intelligence in search, calculation, storage and optimization, and the two kinds of intelligence are highly complementary. A new "1 + 1 > 2" enhanced intelligence, i.e. integrated intelligence, will be formed when people and computers work together to learn from each other, which is a two-way closed-loop system, including both human and machine components. Among them, people can accept the information of the machine, and the machine can read the signal of people. The two interact and promote each other. In this context, the fundamental goal of artificial intelligence has evolved to improve human intelligence and more effectively accompany human beings to complete complex and dynamic intelligent functional tasks.  
Max tegmark, a physicist at Massachusetts Institute of technology, in his book life 3.0: man in the age of artificial intelligence, imagines all kinds of scenes from human evolution to the age of artificial intelligence. After giving his definition of life and intelligence, he divided the development stages of life. He believes that life is a self replicating information processing system, in which information (software) determines its behavior and hardware blueprint, while intelligence represents the ability to achieve complex goals. His division of three stages of life evolution is: Life 1.0, life that hardware and software can only evolve but can't design, such as bacteria; Life 2.0, life that hardware can only evolve, but software can design life to a large extent, which is now human beings, we can't break through the limit of body to run faster than lions, or live longer than turtles, but we can run through our own software All algorithms and knowledge used to process information, as well as the ability to decide what to do, enter our brain by learning and programming the skills we are not born with; life 3.0, not only software can be designed, but also hardware can be designed, which can be the ultimate state of artificial intelligence.


Unlike Stephen Hawking, Bill Gates and musk, who have serious doubts about the AI newspaper, tagmark's attitude tends to be more positive and rational. He believes that the arrival of AI is inevitable, and the following benefits and crises are inevitable, but the benefits are not infinite, and the crises are not uncontrollable. He believes that people should keep a calm and objective attitude towards artificial intelligence. Blind worship and blind resistance will lead to bad results. He also stressed that the change of AI is not only for computer technicians, but also for people from all walks of life including sociologists, economists and other industry experts. AI will affect almost all industries.  
In a word, we advocate that human-computer interaction and coordination can give birth to human-computer integration intelligent form, which should be the core of artificial intelligence 3.0. Through human-computer interaction and collaboration, especially in text, image, voice and other multi-media information to achieve multimedia and cross media interaction and collaboration, artificial intelligence can evolve from a simple computing tool to an indispensable partner of human intelligence. In the future, human-computer interaction will become an autonomous integration within the human-computer environment system, which will organically combine the advantages of human beings with the long board of machines, which is the lifeblood of human-computer integration intelligence. There are many complex and dynamic problems in human society for a long time, which may provide a new solution for human-computer integration intelligence. Therefore, we believe that human-computer integration intelligence is probably the advanced stage of artificial intelligence, and even the final stage. Specifically, human's unique perception (beyond human's rationality and sensibility) can flexibly and freely empower machines and solve the internal paradox of complex dynamic problems. The key elements of human-computer integration intelligence include the balance of yin and Yang of logic and non logic, the balance of yin and Yang of upper consciousness and lower consciousness, the balance of yin and Yang of unity and diversity, and the balance of yin and Yang of knowledge and action.
Intelligent Manufacturing: an important application scenario of AI 3.0
Although AI 3.0 may be used in a wide range of specific scenarios, we believe that one of its most important application scenarios is intelligent manufacturing. In a broad sense, intelligent manufacturing refers to the deep integration of artificial intelligence technology and advanced manufacturing technology, which runs through the optimization and integration of design process, processing and assembly, delivery service and corresponding systems in the whole value chain, aiming to continuously improve the product quality, efficiency and service level of enterprises, and promote the innovation, green, coordination, opening and sharing development of manufacturing industry. In recent decades, intelligent manufacturing has formed many different paradigms in practice, including lean manufacturing, flexible manufacturing, concurrent engineering, agile manufacturing, digital manufacturing, computer integrated manufacturing, networked manufacturing, cloud manufacturing, intelligent manufacturing, etc., which play an active role in guiding manufacturing technology upgrading. It must be pointed out that the current stage of intelligent manufacturing is only the initial stage. In addition, Chinese enterprises have a narrow understanding of intelligent manufacturing, only see the intelligent processing and assembly, and ignore the other two aspects (i.e. design process and delivery service), especially the design process, which is the biggest short board caused by the inherent deficiencies of Chinese manufacturing industry.  
In the face of the emerging new technologies, concepts and models of intelligent manufacturing, it is necessary to summarize and extract the basic paradigm. Generally speaking, the first industrial revolution with steam engine as the core technology is industrial 1.0; the second industrial revolution with motor as the core technology is industrial 2.0; the third industrial revolution with computer as the core technology is industrial 3.0; our unique point of view is that industrial 4.0 is the fourth industrial revolution with man-machine integration as the core technology (instead of replacing human beings with machines)! Therefore, we believe that the current stage of intelligent manufacturing is only the initial stage of long-term intelligent manufacturing, so it can only be industry 3.5, not real industry 4.0. Specifically, the development of industry 3.0 ~ 3.5 characterized by preliminary informatization can be divided into three stages: (1) from the middle of the last century to the middle of the 1990s, informatization is characterized by the primary digital stage characterized by computing, communication and control applications (belonging to the industrial 3.0 stage); (2) Since the mid-1990s, the Internet has been widely used, and informatization has entered the preliminary stage of "ten thousand Internet / Internet of things" (belonging to the primary stage of industry 3.5) characterized by the interaction between people and things; (3) At present, on the basis of cluster breakthrough and integrated application of big data, cloud computing, mobile Internet, 10000 Internet / Internet of things / industrial Internet, AI has achieved strategic breakthrough, and informatization has entered the preliminary intelligent stage (belonging to the advanced stage of industry 3.5) characterized by the new generation of AI technology. Based on the relevant paradigms of intelligent manufacturing and the characteristics of integration of informatization and manufacturing industry in different stages, we can summarize three basic paradigms and stages of preliminary intelligent manufacturing (industry 3.0 ~ 3.5): preliminary digitalization, preliminary Internet of 10000 / Internet of things, and preliminary intellectualization. In our opinion, the current thinking of industry 4.0 only involves industry 3.5, which is far from the real industry 4.0. We boldly predict that man-machine integration is the core of real industry 4.0.



In the western developed countries, intelligent manufacturing is a "series" development process. They take decades to fully develop digital manufacturing, then develop digital networked manufacturing, and then move to a more advanced stage of intelligent manufacturing. In China, the manufacturing industry has a very strong demand for intelligent upgrading, and in recent years, the technological progress is also very fast. But generally speaking, the foundation of Intelligent Manufacturing in China is very weak. Most enterprises, especially the majority of small and medium-sized enterprises, have not completed the preliminary digital manufacturing transformation. Facing this reality, how can China promote the technological transformation and intelligent upgrading of manufacturing industry? We believe that in the process of promoting intelligent upgrading, Chinese enterprises should steadily complete preliminary digital "make-up lessons" (i.e. complete industry 3.0) and lay a solid foundation for the development of intelligent manufacturing, but they can explore a new way of leapfrog development of intelligent manufacturing without following the path of "sequential development" in western developed countries. In recent years, China's manufacturing industry has vigorously promoted the "Internet plus manufacturing". On the one hand, a number of enterprises with a good foundation of preliminary digital manufacturing have successfully transformed and achieved the preliminary ten thousand Internet / Internet of things manufacturing; on the other hand, some enterprises that have not yet achieved the preliminary digital manufacturing have adopted the technical route of parallel promotion of the preliminary ten thousand Internet / Internet of things manufacturing, and have successfully realized the transition to the preliminary ten thousand while completing the "supplementary course" of the preliminary digital manufacturing The leap of Internet of things manufacturing.  
Therefore, China's promotion of intelligent manufacturing can take the "parallel" development route: parallel promotion of digital manufacturing, digital networked manufacturing, a new generation of intelligent manufacturing, and timely full application of high-speed development of advanced information technology and advanced manufacturing technology integration technology innovation, leading and promoting the intelligent transformation of China's manufacturing industry. China will take China Internet as a manufacturing industry in the next few years, taking into account the current situation of intelligent manufacturing development in China. At the same time, considering the immature manufacturing technology of new generation, the key point of the transformation and upgrading of China's manufacturing industry is to promote and apply the Internet plus manufacturing in a large-scale manner. At the same time, we should pay special attention to the integration and application of various advanced technologies in the process of vigorously promoting "Internet plus manufacturing". On the one hand, it enables the majority of enterprises to complete the "preliminary digital supplementary course" with high quality; on the other hand, it makes good use of the new generation of intelligent manufacturing technology as soon as possible, greatly accelerating the transformation and upgrading of the manufacturing industry. In a few years, after the new generation of intelligent manufacturing technology is basically mature, China's manufacturing industry will enter a new stage of comprehensive promotion and application of the new generation of intelligent manufacturing, that is, the long-term focus of the transformation and upgrading of China's manufacturing industry. In the process of transformation and upgrading of Chinese manufacturing industry, German enterprises can effectively help Chinese enterprises to achieve short-term and long-term goals, especially in the aspect of intelligent design process (which is the biggest short board caused by the congenital deficiency of Chinese manufacturing industry).
Organization 3.0: balance of yin and Yang between self-organization and other organizations
 
According to the Enlightenment of tagmark on the definition of life and intelligence and the division of development stages, and with reference to Drucker's discussion on the role of knowledge in productivity, Drucker has seen a fulcrum of organization in the era of artificial intelligence (strictly speaking, it should be the era of man-machine integration Intelligence), that is, the function of organization is the creative output and application of knowledge. In this paper, we put forward a preliminary idea about the characteristics and principles of organizational management in the era of man-machine integration intelligence.  
We believe that organization is a kind of self replicating knowledge processing system, in which knowledge (software) determines the blueprint of its hardware (structure, form) and the specific behavior based on it. According to the role of knowledge in different stages of organizational history, organizations can be divided into three major stages, which are mainly composed of three forms:
Organization 1.0 is an organization form that hardware and software can only evolve naturally, but not be designed. At this time, knowledge only exists in people's mind relatively implicitly, especially the knowledge about organizations has no obvious effect on the structure and form of organizations. Organizations before the industrial revolution belong to organization 1.0. People's roles and functions in the organization are relatively random and scattered. People's understanding of the organization is chaotic, and their attitude to the organization is blind and helpless. At this time, the basic organizational form is a low-level "self-organization" structure (the metaphor of which is "water type" organization).
Organization 2.0 is an organization form that hardware can only evolve naturally, but software can design to a great extent, that is, knowledge can play a great role in the organization itself. After the industrial revolution, the rise of knowledge economy, scientific management, project management, modern management and other organizational knowledge can in turn reshape the structure and form of the organization, but limited by the relative lack of knowledge and the limited rationality of people. At this time, the basic organizational form is a high-level "other organization" structure, i.e. hierarchical organization (the metaphor of which is "mountain type" organization).
Tissue 3.0 is a kind of tissue form that not only software can be designed, but also hardware can be designed. This is the organizational form of "the age of human-computer integration intelligence". Its organizational mission is that everything comes from knowledge, and everything also returns to knowledge. At this time, the basic organizational form is the yin-yang balance between the high-level "self-organization" structure and the high-level "other organization" structure.



In organization 1.0, people's knowledge is limited and not systematic. People are completely subject to the organization, and the power of the organization is weak. In organization 2.0, people begin to have systematic knowledge to act on the organization, and the organization becomes strong. The power of the company has completely changed the face of the human organization on the earth, understanding and transforming the world with unprecedented efficiency. In organization 3.0, because knowledge acts on knowledge, human-computer integration may obtain exponential explosive capability improvement beyond people's expectation. From the organization that mainly maintains the relationship between people and things, to the organization that mainly maintains the interaction and cooperation between human intelligence and artificial intelligence. Artificial intelligence is mainly responsible for the creation and application of knowledge, while human intelligence is mainly responsible for the wisdom beyond knowledge, including the creation and application of wisdom. Organization 3.0 may be the ultimate state of organization in the era of human-computer integration intelligence. Through the efforts of self-organization, human beings make themselves have the objective artificial intelligence created by their own subjectivity. In turn, artificial intelligence can liberate human beings from the primary intelligence work and concentrate on the development of the unique advanced intelligence work, that is, the construction of intelligence savvy. On the one hand, leverage effect is where organizations play an important role. It can be seen that the relationship between AI and human intelligence is based on division of labor and cooperation, supplemented by partial substitution, forming a balance between yin and Yang. The feature of organization 3.0 architecture is to give full play to the great intelligent role of human intelligence. Intelligence refers to the balance of yin and Yang between knowledge and ignorance, and its basis is human-computer integration intelligence. It should be pointed out that supervised learning and autonomous learning of intelligent machines are also the reflection of the balance between yin and Yang.
According to the above imagination of the artificial intelligence era, in the rough outline of the future characteristics of the world of artificial intelligence, we can vaguely find that the organization of "human-computer integration intelligence era" will be a new structure and form, from organization 2.0 to 3.0. Whether it is flat organization (such as project organization), flexible organization (such as water organization), or Internet Organization (such as small and micro organizations and platform organizations), it can be regarded as multi-dimensional and multi angle reflection of organizational transformation.  
Inspired by Alibaba's idea of "Zhongtai" and other enterprises' idea of "platform", we can try to build a new organizational structure that matches with human-computer integration intelligence, namely "three organizational structures". The necessity of dividing organizational structure into three parts is to solve two important paradoxes faced by organizations: (1) the paradoxes of "efficiency / cost" (Standardization) and "flexibility / value" (customization); and (2) the paradoxes of "present" and "future". The paradox of "efficiency / cost" and "flexibility / value" refers to that, on the one hand, data intelligence has a high degree of "reusability", especially using highly shared it technology (such as cloud computing capability, artificial intelligence) to carry out completely different businesses (such as cloud computing capability can support the development of smart cars and unmanned supermarkets), so standardization is required; on the other hand The change speed of the market demand is accelerated, and the user's specific requirements are becoming more and more differentiated and personalized, so customization is required. The paradox of "the present" and "the future" refers to that, on the one hand, the high-intensity competition and high dynamics of the current market require enterprises to focus on the "the present", on the other hand, the complexity of the network ecology, the non designability and uncertainty require enterprises to look at the "future", not be misled by the current problems, and cultivate the ability to predict the future.  
Solving the above two paradoxes and conflicts is the basic principle of the "three Taiwan organizational structure". The front desk and the middle desk can and should be separated moderately. Through the establishment of a small front desk with "light assets", the goal of "flexibility / value" of customized customer needs can be achieved; through the establishment of a middle desk with "heavy assets", the goal of "efficiency / cost" of providing flexible combination of standard modular services for the front desk can be achieved. Front / middle and back can and should be separated moderately. Let the front / middle stage focus on the "immediate" to meet the current changing market demand, while let the independent background focus on the "future" and focus on basic research and strategic exploration. From this perspective, there is a certain degree of paradox conflict between the front desk and the middle desk, that is, the tension between "standardization" and "customization" at the same time and in the same place; there is a higher degree of paradox conflict between the back desk and the middle desk, that is, the tension between the immediate goal and the future goal at different times and in different places.



It seems that the distinction between the front, middle and back three platforms can be based on two specific dimensions: one is the time (present and future) dimension, and the other is the space (distance between customers). However, since the above two specific dimensions are basically the same in direction (that is, the foreground pays attention to the present and is also the closest to the customer; the background pays attention to the future and is also the farthest away from the customer, while the middle and Taiwan are both in the middle), we can combine the two dimensions of time and space into a super dimension, namely agility, That is, reaction speed and reaction degree to external changes. Generally speaking, the agility of the front desk is the highest, that of the middle desk is the second, and that of the back desk is the lowest. According to the principle of "three organizational structures", the basic path of organizational restructuring is "front-end backward moving" (any "front-end" business content that can be standardized, modularized, and reused by multiple "front-end" can be moved back to the middle stage), and "back-end forward moving" (but any business content that needs to cooperate with the front-end, such as the commercialization of basic research and development, and general leadership Training different specific business training, as well as market short-term forecast research different from market long-term trend research, etc., moving forward to Zhongtai).  
Specifically speaking, first of all, the unit size is small, but the number is mostly the characteristics of the front desk (i.e. small front-end), which is the business department with the highest agility requirements, especially close contact with immediate customers and serving immediate customers. The front desk is a highly self-organized business unit, which may become a new elite champion or invisible champion. From the perspective of innovation, the front desk is mainly responsible for quick iterative trial and error of products / services. The metaphor that has a profound significance in this regard is the "water type" organizational form.
Secondly, the service platform with large unit scale is the characteristic of Zhongtai (i.e. large platform), which is the tactical application department whose agility requires to be located in the middle, especially the "platform" service that connects many business departments to provide unified supporting services for immediate customers, so as to realize the transparent flow of data among various business departments of the enterprise. Zhongtai not only endows the internal front desk (such as Taobao), but also endows the ecosystem symbiotic enterprises (such as Taobao sellers) and other customers. From the perspective of innovation, Zhongtai is mainly responsible for the timely application and continuous improvement of products / services. The metaphor that has a profound significance in this regard is the "mountain type" organizational form.  
Finally, the unit size is not large and the number is not large, which is the characteristics of the background (i.e. rich ecology). It is the strategic design, basic research and enterprise logistics department with the lowest agility requirements, and also the coordinator of the future ecosystem (including external partners such as universities, government research institutes, etc.). From the perspective of innovation, the backstage is mainly responsible for exploratory R &amp; D of breakthrough products / services, including business trend analysis and long-term market forecast. The metaphor that has a profound significance in this regard is the "landscape dependent" organizational form.  
In general, "three organizational structures" may be a new organizational structure matching with human-computer integration intelligence. The metaphor of the front stage is a flexible and free "water type" organization form; the metaphor of the middle stage is a stable and solid "mountain type" organization form; the metaphor of the back stage is a flexible and free and stable and solid "landscape dependent" organization form. The future organizational structure may be the balance of yin and Yang formed by "mountain" and "water".
Management principles: Enlightenment in the future
Everything is a unity of opposites of yin and Yang, and everything has two sides of each other. Artificial intelligence is no exception. It is this kind of relationship that promotes the continuous development of things. Human beings endow machines with intelligence, which in turn feeds human beings and promotes the further development of human intelligence. The relationship between human intelligence and artificial intelligence will occupy the core position in a long period of human development in the future.  
The breakthrough development of artificial intelligence will come from the leap of physical limit and the full release of human intelligence, which will involve the revolutionary development of natural science and humanities. Different from the previous stage of human development, this leap of human development may go through a relatively long process. As a functional organ of society, management will still play an irreplaceable role. In the era of human-computer integration intelligence, the role of organization and management may be more critical, because the object of management has been upgraded to a machine with extremely powerful functions and certain human intelligence, as well as organizations and individuals equipped with powerful capabilities by intelligent machines, and the management environment has thus been transformed into a complex giant system formed by intelligent human-computer interaction. Some general management principles need us to conceive as early as possible and take precautions as early as possible.


Principle of assistance
As early as 1980, Carnegie Mellon University made an expert system for DEC company, which can help DEC company save about 40 million dollars every year, especially in decision-making. It can be seen that the history of AI assisted human decision-making has been started for a long time.  
Artificial intelligence does have unique advantages in decision-making. According to the classic decision theory of Amason, the effectiveness of decision depends on two factors: the cognitive ability of decision maker and the interaction process of decision team. Artificial intelligence can provide more comprehensive, accurate and objective information (but based on a certain combination of algorithms and algorithms, the rules are still determined by people), which makes the decision more accurate and efficient. In a system with strict rules, artificial intelligence can make decisions superior to human beings, such as alphago's performance of crushing human beings in go. But in the real world management scene - whether it is the past, the present and the future "man-machine integration intelligence era" - it is almost impossible to build a complete rule system for the real scene which is difficult to exhaust its complexity. Therefore, the value of artificial intelligence lies in the holographic analysis in the relatively complete simulation system to help mankind to propose the optimal decision options.  
The negative influence of AI on decision-making lies in how to coordinate the relationship between human and machine (human error, human emotional factors; resistance to the acquisition of AI information, and measures taken), human privacy, and the implementation status of human under the instruction of AI. Therefore, the first principle of management in the era of man-machine integration intelligence is assistant principle. Artificial intelligence assists human intelligence in organization management decision-making and implementation, rather than decision-making. This is the interactive function between artificial intelligence and human intelligence in the human-computer fusion intelligence as the unity of yin and Yang.  
people oriented
As the unity of opposites of yin and Yang, the artificial intelligence in human-computer fusion intelligence and human intelligence also have the opposite side. "Tao Te Ching" has a saying that "the way of heaven is beneficial without harming, and the way of saints is not contested". Can AI achieve "benefit without harm, for without dispute"? No matter what the answer is, we must strictly guard against the risks that AI may bring to human beings from all levels of design. This risk is not only from AI itself, but also from the damage that some people bring by using AI.  
In the era of human-computer integration intelligence, organizations tend to have an organic balance of "three platforms" interaction, while the information system linking the "three platforms" organization and the artificial control center of each node will constitute a super complex system. It is very important to set up safety valve in each node, and build a whole set of safety control network to ensure that human interests are not damaged. Once the red line of human interests security is touched, the security alarm will be triggered.
Integration of Arts and Sciences
    
In the era of man-machine integration and intelligence, the ultimate management in Drucker's eyes, that is, Boya art tradition, may be truly realized. Because "liberal arts" of liberal arts refers to the comprehensive contents of objective knowledge, subjective wisdom, ethics and Leadership (like "knowledge" in the "unity of knowledge and practice"), while "Arts" refers to flexible and innovative management practice and application (like "action" in the "unity of knowledge and practice"). Managers who practice liberal arts should make use of the interdisciplinary knowledge and insights covered by humanities and natural sciences such as psychology, philosophy, economics, history, ethics and physics, and also focus on the effectiveness and results of knowledge, such as treating patients, teaching students, building bridges, designing and selling products and services. In the era of man-machine integration and intelligence, new requirements are put forward for the integration of knowledge and practice. This puts forward a new challenge to the talent cultivation in the future.  
Modern education originated in Germany in the 17th century. Its purpose is to serve the industrialized production of large machines. What it trains is a uniform labor force. The personality and growth route of students are framed to grow up, which is the "standard talents". However, it can be determined that in the era of "human-computer integration intelligence", the workers and middle-level white-collar workers engaged in repetitive or mechanical work will be replaced by artificial intelligence, while the work requiring human initiative, enthusiasm and creativity and human emotional care needs more human intelligence. Therefore, it can be concluded that future education should be exploratory rather than exam oriented The "unity of knowledge and practice" of creative innovation, rather than simple knowledge accumulation. The most important thing in future education is artistic accomplishment and innovative thinking. In this regard, Ma Yun said very directly: "the next 30 years is the best overtaking era. If we continue the previous teaching methods, memorize, memorize and calculate these things for our children, and do not let them experience them, and do not let them try to play the piano, chess, calligraphy and painting, I can guarantee that the children will not find a job in 30 years."