In the post-epidemic era, which areas will be transformed by artificial intelligence | Roundtable
- Categories:会议新闻
- Author:
- Origin:
- Time of issue:2020-12-24 08:43
- Views:
(Summary description)The 2020 outbreak of New Crown Pneumonia poses an unprecedented challenge to the Chinese economy. In response to the economic impact of the outbreak and with an eye on China's future economic development
In the post-epidemic era, which areas will be transformed by artificial intelligence | Roundtable
(Summary description)The 2020 outbreak of New Crown Pneumonia poses an unprecedented challenge to the Chinese economy. In response to the economic impact of the outbreak and with an eye on China's future economic development
- Categories:会议新闻
- Author:
- Origin:
- Time of issue:2020-12-24 08:43
- Views:
The 2020 outbreak of New Crown Pneumonia poses an unprecedented challenge to the Chinese economy. In response to the economic impact of the outbreak and with an eye on China's future economic development, the Chinese government has launched a new round of infrastructure investment. Unlike in the past, this round of infrastructure investment is focused on new technologies for the future and further complementing the current infrastructure. The new infrastructure has been described as the "new engine of China's economy in the post-epidemic era". The development of new infrastructure is closely related to the development of artificial intelligence. At the 4th Global Entrepreneur Conference, a concurrent conference at the Shenzhen (International) Artificial Intelligence Exhibition on August 28th, corporate leaders from the AI field sat down for a roundtable discussion on the development opportunities and changes brought about by the new infrastructure.
The forum focused on the theme of "What will AI bring changes in the post-epidemic era", with the following main discussions.
1. What kind of opportunities and challenges the epidemic brings to AI companies
2. How to use the "new infrastructure" to better help develop business
3. Under the favorable policy of "new infrastructure", in which direction do AI companies need to breakthrough in the future
4. Under the new situation, what other fields do you want to enter, and what are the industries with the greatest opportunities?
Roundtable Guests.
Xiao Jing, Chief Scientist of Ping An Group and Rotating Chairman of Shenzhen Artificial Intelligence Industry Association
Zheng Wenxian, Vice President of Shenzhen YuntianliFei Technology Co.
Lei Hu, Founder and CEO of Beijing Zilu Deep Vision Technology Co.
Jia Yuhang, General Manager of Beijing Cloud Test Information Technology Co.
Fu Qiang, Head of Commercialization Products, Shanghai Heavy Union Information Technology Co.
Note: The forum was hosted and questioned by Xiao Jing, Chief Scientist of Ping An Group and Rotating President of Shenzhen Artificial Intelligence Industry Association.
01
What kind of opportunities and challenges does the epidemic bring to AI companies
(Xiao Jing, Chief Scientist of Ping An Group and Rotating President of Shenzhen Artificial Intelligence Industry Association)
Xiao Jing: Last year there was a statement that 2019 is the worst year in the past ten years, may be the best year in the next ten years, when it was thought that may economic downturn may happen, did not expect 2020 is actually so strange, even the meltdown occurred several times, now the epidemic is so serious, caused a great impact on our lives, work also caused a great impact, very good phenomenon many people become more and more affluent, eat very well.
In this case, the epidemic is now more and more serious in many foreign places, China control is still very good, but the epidemic control may also become a normal initiative, there may be a small wave of rebound in the fall, winter, many people are worried about this, this situation, you think artificial intelligence companies will have what kind of challenges or opportunities, on this issue, you tell us their views.
Zheng Wenxian: from the cloud sky this half-year experience, the outbreak of the epidemic is very sudden, testing the ability of artificial intelligence companies, or technology companies to adapt to change. From the third day of the Lunar New Year, the epidemic situation is very serious, began to control the flow of people, I also went with the technical team to Shenzhen North Station, to the core high-speed intersection to see the card, card points, the deployment of some AI face temperature measuring instrument, as well as how to use the original system has been built in Shenzhen, better for these epidemic prevention personnel management play some role. Such an epidemic scenario brings some opportunities and technical challenges to the enterprise, for example, before doing portrait rarely will face today all people will wear masks, recognition algorithm accuracy and training will be affected, for the algorithm built by the enterprise is bound to adjust themselves.
The second is the ability to launch products and solutions to solve problems. In fact, our system in February, to help the police play some role, for example, in Longgang District, there is a New Year's Day to Wuhan to visit relatives by car, after the occurrence of his return to Shenzhen, but he did not give their community personnel to report, and he may go to other areas many times, this is very dangerous, but through our epidemic prevention system to assist the police and community epidemic prevention personnel to quickly screen out his close But through our epidemic prevention system, we assisted the police and community epidemic prevention personnel to quickly screen out his close trajectory, a close contact, about 75 people, and then quickly minimize the impact, which is the role played by this type of artificial intelligence enterprise in this epidemic.
At the end of February, Wu Manqing, an academician at the State Council's joint prevention and control press conference, also praised Yuntian Lifei's products and solutions.
In terms of challenges, we will find that people are very concerned about data security and privacy. Whether it's travel codes, or related kinds of prevention codes, or portrait-based solutions, the public is more strongly concerned this time, and I believe this is also the process of cultivating the public. For practitioners, now they need to actively face with the government how to do data integration, how to do data security, which puts forward a new requirement and a new challenge for us.
I think there must be opportunities in the midst of challenges or problems, and I always talk to my team to see opportunities in the crisis. I have always told my team to see opportunities in crises. In the context of this epidemic, which is a "black swan" event with significant global impact, we have indeed seen some opportunities. I heard a very interesting thing some time ago. In the past few years, in the summer, all children's hospitals were overcrowded with children, because the spread of HFMD was very serious. This year, the children's hospital is empty, because everyone wears masks, forming a very good hygiene habits will not spread, this may be the positive things brought by the epidemic.
For our own 3D-related field, we also see a good opportunity. It may often encounter problems before the need to upgrade the technical means above, to the development of new technologies to bring the role of push. To cite a few examples, we found that when people wear a mask, perhaps because the lower half of the face is obscured, in this case face recognition accuracy will be more affected, there is also the image angle, can also be targeted experiments, but less features this thing is physical, is the data research, it is difficult to change, itself 3D because of the shape information latitude supplement, still can get a good enough features, can effectively help to face recognition in the case of wearing a mask.
In fact, we have made some attempts before the epidemic, because in the security requirements, we have done some work for intentionally taking scarves or masks to cover people's faces, and further strengthened this piece during the epidemic, and found that this piece is very effective.
During the epidemic, in order to prevent the epidemic, in order to track potential contacts or transmitters of the epidemic, the resident population, including all people need to carry out relatively fine control, of course, visual is only one aspect of the control method, just mentioned a lot of big data technology and a variety of technologies have played a very important role in it, in turn, in fact, also promote the public for such new technology experience and feelings, more easily accept it.
There are also non-contact, including in the smart home door lock, as well as many office or industrial application scenarios, we further upgrade the imaging face or other non-contact biometric way to achieve the door unlock, or other applications, we feel very logical, or should go to the leading direction, these are some of the opportunities brought to us.
Fu Qiang: Hello everyone, first of all, a brief introduction to ZM Technology, ZM Technology was founded in 2013, is a focus on business intelligence technology services platform, we have been focusing on traditional enterprises in efficiency and cost reduction, as well as the core proposition of risk prevention and control.
The first thing we see is that the epidemic has a great impact and influence on the lifestyle of the population. We should feel it deeply, from spending more time offline before, gradually shift to more online, including outdoor actions, behavior has also been some restrictions, and gradually shift to activities at home.
In fact, these have a great direct impact on our consumers in the consumption scene, consumption judgment, can be expressed as the transformation of online consumer purchases.
Secondly, we have also observed that the specific structure of the industry also has a considerable impact, as we can see that the epidemic has caused a greater impact on more traditional labor-intensive enterprises that may focus on offline service scenarios.
Based on the above background, we can see that AI is in fact an intelligent upgrade of the whole society, and I think after this epidemic, there should be more accelerated development trend.
First, in the epidemic prevention-related fields, for example, just introduced like security, travel, and also including the field of disease detection, there is a very rapid application, also in the epidemic prevention China can achieve such good results, played a very important role.
In this process, Zhongmeng Technology also quickly responded to the risk prevention and control, including offline big data tracking technology applied to the community, through some WeChat small programs lightweight rapid iteration, to many offline public security, community scenes, etc. to provide support, although the point is not very large, but you can see the epidemic in all aspects of society on the application of artificial intelligence technology acceptance has grown significantly.
As the epidemic begins to ease and the economy resumes its growth phase, and as the "new infrastructure" policy continues to take hold, we believe that AI technology should continue to accelerate in more industry sectors.
In addition, in the more traditional and lower areas of business, will think that it actually brings greater opportunities, the above are our views on this piece.
Jia Yuhang: Hello, I'm Jia Yuhang from Cloud Data. Briefly introduce cloud measurement data, cloud measurement data is to provide high-quality scenario-based AI data services for artificial intelligence enterprises, combined with the theme just mentioned, I actually think that artificial intelligence enterprises in the current epidemic of opportunities and challenges coexist.
First of all, combined with the current characteristics of artificial intelligence and deep learning, in the depth of the use of scenarios, interaction mode depth binding, the arrival of the epidemic has greatly changed the way people travel, interaction mode. For artificial intelligence products how to bring new interaction mode in the new epidemic environment, new travel mode scene design, problem design and training data related to a series of issues, it will actually face a new reconstruction. For example, online education, live broadcast and other industries in all walks of life after this year's epidemic is gradually heating up, more and more different industries will also be tools online, electronic, and thus generate a lot of data, these data can be used for intelligent, help in the mobile office, intelligent era of new development of enterprises.
02
How to use the "new infrastructure" to better help develop business
Xiao Jing: All of you are full of confidence, it feels like you should adjust the KPI of the company, if the market value does not double or triple by the end of this year, it should be considered not to meet the target, and the CEO should be replaced.
This year, in addition to the epidemic, there is a big thing, that is, the "new infrastructure". Just now Mr. Zheng also mentioned a lot, the country has increased a lot of investment in 5G, Internet of Things, data centers, etc. The construction of the "new infrastructure" will certainly bring a lot of benefits to the entire AI industry, including AI artificial intelligence field enterprises, will also bring a lot of new opportunities, there are also a lot of points to make full use of and fully participate in, please talk about your Please talk about your own enterprises, under the big situation of "new infrastructure", how do you layout, how do you consider, not only conceptually speaking, what will help us, infrastructure construction, how do you put into the "new infrastructure" construction of the big environment, how to use the "new infrastructure" construction. How to use the "new infrastructure" infrastructure capabilities to help you better develop your business.
(Zheng Wenxian, Vice President of Shenzhen Yuntian Lifei Technology Co.
Zheng Wenxian: Under the general direction of "new infrastructure", Cloudsky has made some layout by combining its own accumulation in algorithm, chip and data platform. The data center is more functional in computing, or in storage, but the real future data center may have to be placed more in front of the application layer, so that the real computing platform can better provide corresponding services for the industry, because in the past the supercomputing center and industry relevance is not very high, so in the next stage, the next phase. Under the wave of "new infrastructure", with the continuous development of artificial intelligence technology, we also want to further promote the construction of artificial intelligence computing enabling platform.
The "new infrastructure" underlying hardware base Cloudy is now involved, another advantage of Cloudy is that we have sought a number of algorithm bins on top of the solution base for so many years, adding the algorithm bins to the hardware layer, packaging hardware + software independent and controllable, based on the Xinchuang computing empowerment platform, is now also in cooperation with many domestic parks and local governments to build a new infrastructure for application solutions.
(Lei Hu, Founder and CEO of Beijing Dulu Deep Vision Technology Co.
Household Lei: I think the word "new infrastructure" is essentially an infrastructure with certain new ideas and new technology applications, and it is still essentially an infrastructure, but the infrastructure may not be entirely in the original physical space, it may be information infrastructure, or data infrastructure, intelligent infrastructure will also be integrated into this concept, all the companies doing AI, what they are doing is participating in the "new infrastructure".
My personal view on the "new infrastructure" is that I think a complex system, regardless of whether the business is complex or mechanically complex, I think it is often more fragile, but simple is easy to expand, and can really stand up and generate value.
So I think the "new infrastructure" may end up in the macro beyond the big things, and may end up in a very simple breakthrough in the scenario of the point.
For example, for example, the construction of rail transit, airports, etc. is now significantly promoted, which through intelligent capabilities and information means, including data collection and some analysis and application of the above can make the application more efficient, including now involved in the rail transit like face ticket entry, in fact, the ticket itself is an infrastructure, rail transit is also a type of infrastructure, which is also constantly under construction, to facilitate travel and life This is within the scope of the ability to participate in things, including some of the smart home experience, it can eventually enter thousands of households, whether from the consumer sector, or in the B-side of some areas, because there is a typical rental housing or shared housing management needs, accurate management of the relationship between people, rental housing or shared housing infrastructure laid in, which has both identity recognition, and 5G, and a variety of technologies in the above. A variety of technologies in the above, big data some analysis, these are aspects of the construction of 5G infrastructure.
Fu Qiang: The core of the "new infrastructure" or the use of some of the most advanced technologies and capabilities to help society in all aspects of construction to accelerate, so as to promote the entire society to intelligent perspective on the upgrade.
For first- and second-tier cities, such as Shenzhen, smart city construction, intelligent driving and other new technologies, new areas have actually been applied more in-depth and extensive. But this time, affected by the epidemic, in the traditional sinking market, their understanding of "new infrastructure" and digital transformation is accelerating. This is consistent with the general direction of helping enterprises to carry out digital transformation, because they need to use more advanced technology to empower and enhance. The core focus of Zhongmeng Technology is still on the sinking market, using intelligent marketing technology to help enterprises achieve digital transformation and upgrade.
(Jia Yuhang, General Manager of Beijing Cloud Measurement Information Technology Co.)
Jia Yuhang: I think in fact AI as one of the core elements of the "new infrastructure", in fact it and several other "new infrastructure" elements have some common points. Like 5G, ultra-high voltage, new energy, etc., in fact, more based on the concept of pipeline transmission, in fact, artificial intelligence also has such a pipeline, transmission characteristics, as mentioned in Zheng's speech, just now, in fact, artificial intelligence gradually transformed the physical world into data assets, such a process in fact, artificial intelligence plays a role in transforming reality into the virtual world digital twin technology means, as As for the data algorithm, computing power and data service providers behind such technology as artificial intelligence, they actually attach great importance to the use scenarios and data processing methods corresponding to the "new infrastructure" under the development of 5G and the "new infrastructure", which should be the corresponding power point of data. Based on the "new infrastructure" development direction, help enterprises to completely restore the real scenario of the corresponding processing scenario, and in this corresponding scenario to complete the data collection work, data processing work, to help enterprises can quickly and efficiently obtain high-quality training data, to help artificial intelligence enterprises in the fastest time to complete the algorithm iteration, and can be applied and practiced in the The company is able to apply and practice in the society.
03
In the new infrastructure favorable policy, artificial intelligence enterprises in the future in which direction need to breakthrough
Xiao Jing: I think artificial intelligence enterprises should rejoice, "new infrastructure" construction is a very good thing for everyone, infrastructure is very costly, the Internet of things for so many years, why has not been up, a big reason, is really to build the Internet system is very costly, general small businesses which have so much money to build so many sensors everywhere to do.
Now the state is willing to help everyone to do a good job on this infrastructure, 5G, data, intelligent computing center, supercomputing center, to help you do a good job on these, which can be solved as long as the specific problems, so artificial intelligence companies should rejoice, very happy, solve the important pain points, because artificial intelligence on the ground is actually very difficult, there are many challenges, one of the challenges is the pre-investment is very large, high human costs, a variety of computing costs, and other facilities sensor costs are very high, so someone to help you solve such a pain point should be very happy, very happy thing.
The country helped us to find a way to solve the infrastructure, what else do you think in the process of artificial intelligence implementation, you feel that in your actual work encountered real challenges, you think the future also need to spend what energy breakthroughs, in your own actual work, what problems encountered in the application field.
Zheng Wenxian: It is true that in the past few years, there are many challenges in the process of implementation and promotion of the scene, in fact, for example, one is the public, or the government, including the community's expectations of artificial intelligence, I think it is relatively high, or for artificial intelligence technology to change the world's expectations and ideas, expectations are relatively high, but indeed as a practitioner, artificial intelligence, whether from the technology or industrial implementation process There is still a long way to go, just mentioned the algorithm, now is the establishment of the deep learning framework, will there be a new framework proposed, to the maturity of the algorithm can meet the future 5G era of social life in all aspects, our computing power is not more lightweight, low-cost efficient bearing algorithm, can really realize the smart connection of everything.
In fact, we may have very good expectations for 5G, artificial intelligence, or very eager for a better life, but this will bring new momentum and challenges for practitioners, indeed we also hope that we can be more calm and rational view of the trend of such technology development, do not need to overheat to overnight to the artificial intelligence technology industry empowerment labeled some, this may be a recent I often and whether it is investors, or and the government to communicate with some kinds of customers a point.
Household Lei: I think artificial intelligence is a technology, in fact, the most difficult to its implementation or to combine with the industry, so based on the entire industry application demand scenarios have a very deep understanding and understanding can do very well, because the technology itself into a scene can serve the industry, or into a solution, into a scene to serve the industry or scene.
I have heard a saying, the product is the carrier of imagination, I for technology and products bearing imagination should be and industry and the actual needs to be able to fit well, before you can embed it, and it can not bring too much burden to the industry, can not bring too much cost, because the cost is not only the cost of the product itself, but also includes the cost of the application, now still see the problem in the middle, many times need to industry The business itself, or a variety of architecture, including responsibility for the demarcation of combing or institutional policy changes, or adaptive technology to be able to, which is between the two closer to each other, to quickly solve the problem of the last mile.
Both the need for AI technology companies to work hard, but also the need for industry and society to have a more inclusive and welcoming attitude to the new technology, to use it up.
(Fu Qiang, Head of Commercialization Products, Shanghai Heavy Union Information Technology Co.)
Fu Qiang: Regarding the challenges, we think the biggest challenge is who can first put simple and direct, especially low-threshold applications to the actual implementation of various industries. Combined with the impact of the epidemic, the next biggest challenge is how to be faster, lower threshold, simple and direct application of technology to all walks of life, because the technology really has to be combined with the ground, the actual application can bring out the value, for the focus of attention like the traditional industry digital and transformation direction, the vast majority of traditional Chinese enterprises, some of their business models, including thinking mode, or They are still stuck in relatively traditional thinking. This may also be the direction that CML Technology has always wanted to try and explore.
Jia Yuhang: I very much agree with the combination of artificial intelligence and industry, in fact, artificial intelligence as a technology will eventually be used as a product by the user, in fact, artificial intelligence has been from the original only to do recognition, gradually from the perception of intelligence to the level of cognitive intelligence, the actual technology research and development process, has been more than a single aspect of computer science, but the combination with the industrial level, the academic community, industry, too. The same, simply take autonomous driving as an example, maybe three or four years ago to do the labeling of autonomous driving-related environmental perception, only need to label out this is a person, this is a car, but now, we need to distinguish whether this is a two-box car or a three-box car, Wuling Hongguang in the end is a car or van different segmentation scenarios.
In order to solve the gradually increasing data industry threshold, the industry gradually because of the depth and industry combination of the industry threshold, in fact, on the one hand, through the accumulation of self-built data labeling base, gradually let us engage in personnel with domain knowledge, on the other hand, through the technical iterative way to help process personnel gradually reduce the industry cognitive threshold, to help in artificial intelligence gradually in all walks of life to play the role of data behind.
04
Under the new situation, what other fields want to enter, what is the industry with the greatest opportunity
Xiao Jing: very good, now we all have a consensus, artificial intelligence technology to be combined with industry to produce a good effect, I remember five years ago I participated in a similar forum, I said artificial intelligence is difficult to land, must be combined with industry, to respect the business, verify the program, not just rely on technology, only show off technology, at that time recognized by few people, most people are talking about my artificial intelligence technology how stable, high accuracy, how powerful the level, but few people look at the industry, now we have basically formed a consensus, this is very good.
There are many kinds of industries, as Mr. Jia just said, we have a variety of different industries. The news this morning, China's micro-industrial, July profits have increased by 19.6% year-on-year, three consecutive months of growth, so the traditional industry may feel that it is lagging behind, but in fact, the development is still good, in this situation, and finally ask a question, we briefly answer.
You think in this situation, artificial intelligence technology in addition to the current industry of concern, you want to help, for example, like security, etc., or perhaps financial, security, these areas, in addition to these you think the most want to enter, think the industry with the greatest opportunity is what industry?
Zheng Wenxian: is currently making efforts in the social governance industry, is also related to the epidemic, will find like Beijing, Shanghai, Shenzhen, such mega-cities, the rapid growth of the population, but and the level of social governance, in fact, does not match, I believe this piece of artificial intelligence technology will have a lot of space to play.
Household Lei: In fact, we do 3D, I still hope to do something that everyone can use, especially consumers can use things, of course, now can also see this possibility, perhaps within 2-3 years, the real virtual digital people and AR, VR-related applications will gradually come, may change everyone, whether in the entertainment, consumption process, or the process of life in the new transformative, disruptive experience, from a technical point of view, gradually matured, which we think very much want to be able to have our own contribution to the direction.
Fu Qiang: Zhongmeng Technology has always been very focused on intelligent marketing, how to put artificial intelligence technology in marketing, including the effect, because we think there is still a lot of space to be explored, today I want to share with you, is heavy Zhongmeng Technology has spent a lot of resources and time in the construction of offline marketing scene online. As we all know, online media has actually realized online and data tracking, but it may be difficult to quantify the offline scenes of people's daily life through data tracking and measure it well in the marketing field.
But with the smart phone, including offline smart devices continue to popularize, in fact, we also continue to build and improve over the years, including the combination of online data intelligence analysis technology continues to iterate, at present, Zhongmeng technology has been in the coverage of offline marketing scenarios have more coverage results, hope to continue to apply these results to offline traditional enterprises, their marketing target group identification and analysis, this piece is our focus to continue to break through a point.
Jia Yuhang: Cloud measurement data has actually been in the driving, financial, home and smart city industry to do the whole industry data services, in fact, including just may be more horizontal expansion, in fact, we would like to next development direction is to vertical, more in-depth development, that is, from the whole artificial intelligence industry, from the whole perceptual intelligence up to cognitive intelligence need to deal with, like semantic information, including knowledge map construction related to the corresponding content, so that cloud measurement data has accumulated a large number of financial industry, home industry corresponding to the professional talent completely related to knowledge map, including the actual extraction, AR-related construction capabilities.
Xiao Jing: Very good, I look forward to your companies to help traditional industries, including urban governance is getting better and safer, this is all for today's forum, thank you all guests, thank you.
