编程知识 cdmana.com

Academician Li Jie: due to the absence of core algorithms and related system software, artificial intelligence may also face the dilemma of "sticking neck"


China Science and Technology News Network 12 month 21 - ( Nashan )“ In the field of artificial intelligence , Compared with other countries , In terms of core algorithm and related system software, we also have the possibility of neck sticking .” Recently held “2020 Big data technology communication and application Summit Forum ” period , Professor of Shanghai Jiaotong University 、 Li Jie, a foreign academician of the Japanese Academy of engineering, said in an interview with China Science and technology news network .

Lack of core algorithm and related system software , Get caught in “ thetime ” pas

Although the development and application of artificial intelligence is in full swing in China , But the innovation ability of China's AI industry is not strong enough , Industry development also depends on open source code and existing mathematical models , There are not many things that really belong to China .

Especially in the core algorithm and related system software , The development of artificial intelligence is inseparable from algorithms 、 Calculate the force , It can be seen as the core and key point of artificial intelligence . however , Compared with developed countries , There is still a big gap in China .

“ We use foreign algorithms and system software , Its professionalism and pertinence need to be improved . in addition , Because of the high cost , It can not meet the actual requirements of specific tasks .” Professor Li Jie said , such as , Artificial intelligence algorithms developed with open source code can accurately recognize faces , But in the recognition of medical images, it is difficult to meet the clinical requirements .

Our country also has a deep dependence on the system software framework from abroad , It's also a big part of the AI ecosystem “ Short board ”. May slow down 2030 The implementation of the plan to narrow the gap of artificial intelligence technology with advanced countries before 2000 , In China, the basic software and system algorithms related to artificial intelligence need to be further improved .

According to the 《2019 White paper on the development of global artificial intelligence 》 Statistics , Our country has been continuously 6 It became the largest robot application market in the world , China has the largest demand for chips in the world , But many high-end robots are imported from abroad . The gap between the core algorithm and the system software , Is the domestic industrial robot to high-end manufacturing “ Land Rover ”.

If that doesn't change , It is very difficult for the application of artificial intelligence in China to go deep , It's also hard to get significant results .“ Once being ‘ thetime ’ It will be very passive , So the underlying algorithms and related software systems should also be strengthened .” Professor Li Jie mentioned .

The accumulation of big data provides raw materials for the development of artificial intelligence

Professor Li Jie said , Big data is the basic raw material for the development of artificial intelligence . AI makes a lot of data valuable . No data , Or less data 、 Data latitude is less 、 Not detailed enough , Artificial intelligence algorithms don't work .

At present, the deep learning of artificial intelligence is mainly based on big data , That is to train artificial intelligence through big data , And summed up from the computer can be used in similar data on the knowledge or law . Although many people define it as “ Big data is big data ”. however , This statement is not accurate !“ On a large scale ” It's just the amount of data . Large amount of data , It doesn't mean that data must have value that can be used by deep learning algorithm .

for example : As the earth moves around the sun , Record the speed of the earth relative to the sun every second 、 Location , You can get a lot of data . If there is only such data , In fact, there is not much value to tap .

We can refer to Martin · Hilbert's conclusion , Big data is actually 2000 After year , Because of information exchange 、 Information storage 、 Data generated by the rapid growth of information processing capabilities in three areas .

Big data mainly includes collection and preprocessing 、 Storage and management 、 Analysis and processing 、 Visual computing and data security , The scale of data is constantly expanding 、 A wide variety 、 Fast production 、 High processing capacity is required 、 Strong timeliness 、 Reliability requirements are strict 、 High value but low density , Provide rich data accumulation and training resources for artificial intelligence .

Take the number of training images used in face recognition as an example , Training face recognition systems of major companies require a large number of face portraits .

The powerful booster behind AI : Cloud computing

Cloud computing is to provide global user computing power through the Internet 、 Storage service , Provide hardware foundation for Internet information processing .

Big data uses cloud computing technology to obtain valuable information from the vast amount of Internet information for information induction 、 retrieval 、 Integrate , Provide software foundation for Internet information processing .

Cloud computing is the foundation , No cloud computing , Unable to realize big data storage and Computing . Big data is an application , No big data , Cloud computing lacks goals and values . Simply speaking : Cloud computing is the virtualization of hardware resources , Big data is the efficient processing of massive data .

Both big data and cloud computing need artificial intelligence , Artificial intelligence is a commercial application after the Internet information system is orderly . AI is the real outlet of cloud computing and big data .

for instance : The underlying algorithm is equivalent to “ housewife ”, Big data is equivalent to “ rice ”, Cloud computing is equivalent to “ Tools ”. No data , Or the data quality and quantity are not high enough , These algorithms “ housewife ”, It will also face “ No rice in the pot ”, Can't make delicious food . With “ big data ” And cloud computing , We can easily order these algorithms “ housewife ” They give us “ I'm cooking ”.

Artificial intelligence reshapes all walks of life

Professor Li Jie said , In our country , Although compared with foreign countries in the algorithm 、 software system 、 There is a gap in robotics and so on , However, there are many application scenarios in China , Data processing is also relatively perfect , Social governance applications 、 Compared with foreign countries, it's not bad at all . For example, the prevention process of this new outbreak , Big data and artificial intelligence have played a big role .

With 5G/6G The arrival of the , The commercial application of artificial intelligence has become the focus of attention , And it's fast moving forward . For example, we are familiar with portrait recognition 、 Image recognition technology 、 speech recognition 、 natural language understanding 、 User portrait, etc . This kind of technology is now in Finance 、 Internet of things and other industries have gradually been more widely used .

Different industries face different pain points , For example, the financial industry is facing cost pressure 、 Products and services are single 、 Transaction fraud, etc , Both the health care and education industries face unequal resource allocation . Although the problem is different , But through data collection 、 Processing and analysis can effectively solve the various problems mentioned above , And artificial intelligence can well empower related industries and industries through data-driven .

“ Artificial intelligence needs to develop , It is necessary to analyze the characteristics and characteristics of relevant industries or industries , Break one by one , Do a good job in ecology , common development .” Professor Li Jie mentioned , Because a company or a country is stronger , It is also difficult to make all industries and industries stronger and better , To be strong, there must be ecology , Work together to do .

For the future , Artificial intelligence will be in every aspect of human life , Will play more and more roles , In the near future, more and more intelligent industries and industrial application systems will appear , Such as smart health care and education 、 Intelligent driving 、 Smart factory is on its way to here !

This article is from China Science and technology news network

版权声明
本文为[Zhang Youwei]所创,转载请带上原文链接,感谢
https://cdmana.com/2020/12/20201224133945393p.html

Scroll to Top