论文部分内容阅读
Chinese technology and e-commerce giant Alibaba has developed artificial intelligence （AI） software capable of outperforming humans on a rigorous global reading comprehension test， according to the company’s English language news platform Alizila. The breakthrough marks the first time machines have surpassed people on languagebased tests and could have a major impact on the future of employment worldwide.
On January 5， a machine-learning model developed by Alibaba’s AI research branch scored higher than humans on the Stanford Question Answering Dataset （SQuAD）， a credible system for the evaluation of machine reading. The achievement was confirmed in a tweet from one of the test’s creators， Pranav Rajpurkar， on January 11. Alibaba scored 82.44， while the figure for human performance registered by Stanford University stood marginally lower at 82.304. Microsoft achieved a similar feat， with its entry registering a score of 82.65 a day later than Alibaba’s.
Man and machine
Competition between humans and AI is nothing new， with computers long able to challenge people in strategically complex games such as chess. The ancient Chinese board game Go（weiqi）， however， had always been considered a much greater trial for machines to overcome， with more potential board configurations than there are atoms in the universe， at least until Google took up the challenge in 2014. The company’s AlphaGo research project was created to test how well a neural network using deep learning could compete at Go.
Beating its first human opponent in 2015， AlphaGo relied on mechanisms similar to the human brain to improve its playing style from experience， before beating a world champion over five highly publicized rounds in Seoul in 2016. However， according to experts， games such as Go and chess rely more heavily on memory and computing power， which computers have in abundance， while language-based tests of the kind evaluated by SQuAD are far more difficult for AI to contend with.
SQuAD， the system used for the evaluation of Alibaba’s technology， was created by researchers at Stanford University as a means of testing the ability of machine-learning models developed by corporations， individuals and academic institutions. Launched in 2016， it consists of over 100，000 questions on more than 500 Wikipedia articles， where the answer to every question is a segment of text from a reading passage. It is significantly larger than previous reading comprehension datasets and correspondingly more challenging. According to the developers， SQuAD is unique compared with other benchmarks because its means of testing is hidden. Teams submit their code which is run on a test set that is not publicly readable， crucial to preserving the integrity of the test results. The topics of the questions are extremely broad， ranging from 16th-century religious figure Martin Luther to the immune system and from European Union law to prime numbers. One of the selected articles focuses on the Chinese Yuan Dynasty （1271-1368）， throwing out obscure questions such as“what interfered with Kublai Khan’s second invasion of Japan？” that would even have history buffs reaching for a search engine.
The competitiveness of the SQuAD leaderboard has already led to the significant advancement of novel deep learning software， as teams contend to build machine comprehension systems that can outperform those of their rivals. A number of renowned international academic institutions and global technology companies have submitted their models so far， including Microsoft， Tencent， Facebook and Tsinghua University.
Rewriting the future
While the win constitutes a major success for Alibaba and Chinese technology in general， the new systems of AI generated by the project are expected to have wider and potentially profound consequences for employment， particularly in the customer service industry.
According to Luo Si， head scientist for natural language processing at Alibaba’s research institute， machines will now be able to respond to simple questions with a high level of accuracy. “We believe the underlying technology can be gradually applied to numerous applications such as customer service， museum tutorials， and online response to medical inquiries from patients， freeing up human efforts in an unprecedented way，” he was quoted as saying on the Alizila site.
In fact， Alibaba has already been deploying similar technology in real customer service situations. When customers call Alibaba’s helpline with a question， they are usually put through to the internal computer system Ali Xiaomi. Ali Xiaomi deals with spoken and written queries， providing answers to frequently asked questions and specific inquiries pertaining to delivery status， wish lists， and recommendations. The system can automatically access account information for callers， and through the same kind of mechanisms that allowed AlphaGo to teach itself Go， Ali Xiaomi is continuously upgrading its own capabilities and learning through experience.
The technology is efficient and cost effective compared with human customer service， and， according to the company， has been especially useful in managing the huge volume of inbound inquiries during the November 11 Singles’ Day shopping festival in China. It has been so successful， in fact， that Alibaba is offering customizable， textbased versions of the technology free to all merchants on their platforms so that they can automate their own customer service. Based on this level of proliferation， observers are predicting that such technology may soon replace human-operated customer service positions worldwide. AI， the future of its application and development， was among the most talked about topics at the recently concluded World Economic Forum （WEF） annual meeting in Davos， Switzerland. Industry leaders came down on both sides of the debate， positive and negative， about the impact of AI and its associated technology on mankind. Despite his company’s recent success in the field， Jack Ma， founder and Executive Chairman of Alibaba， struck a cautious tone. “AI and big data are threats to human beings. I think AI should support human beings. Technology should always do something that enables people， not disable people，” he said at the WEF.
“The computer will always be smarter than you are； they never forget， they never get angry， but computers can never be as wise as man. AI and robots are going to kill a lot of jobs， because in the future it’ll be done by machines. Service industries offer hope， but they must be done uniquely，” Ma said.
Elsewhere at the forum observers were unanimously positive in their appraisal of AI’s potential， with many quick to point out that historically technology has created more jobs than it has destroyed. Google CEO Sundar Pichai expressed his view that “any time you work with technology， you need to learn to harness the benefits while minimizing the downsides.”
“AI is probably the most important thing humanity has ever worked on. I think of it as something more profound than electricity or fire，” he said， explaining that the technology could eliminate many of the constraints humanity now faces， helping for example to make “clean， cheap， reliable energy” a reality.
A new frontier
Since being founded in Hangzhou in 1999， Alibaba has gone on to become one of the world’s 10 largest companies with a market value of $490 billion， primarily through its successful e-commerce platforms. Toward the end of 2017， the company revealed grandiose plans to invest $15 billion in emerging technologies such as AI， quantum computing， and the Internet of Things， with its leading machine reading technology just one aspect of a much larger research initiative.
The breakthrough arrives amidst a wider climate of technological innovation in China. The World Internet Conference was held in Wuzhen， east China’s Zhejiang Province， in December 2017， where global technology leaders employed powerful rhetoric to herald a new age of consumer technology with AI at its core.
China plans to lead the vanguard of global AI research， development and application according to plans issued by the State Council last July， described by Vice Minister of Science and Technology Wang Zhigang as a major step for carrying out an innovationdriven development strategy and making the country a global leader in science and technology.
The plan stated that AI has the potential to serve as a major driver of economic growth as the shape of the Chinese economy undergoes a major transition toward innovation， capable of tangibly improving people’s lives by 2020. The release also set the goal for China to become a major global center for AI development， leading the world in AI technology by 2030.
On January 5， a machine-learning model developed by Alibaba’s AI research branch scored higher than humans on the Stanford Question Answering Dataset （SQuAD）， a credible system for the evaluation of machine reading. The achievement was confirmed in a tweet from one of the test’s creators， Pranav Rajpurkar， on January 11. Alibaba scored 82.44， while the figure for human performance registered by Stanford University stood marginally lower at 82.304. Microsoft achieved a similar feat， with its entry registering a score of 82.65 a day later than Alibaba’s.
Man and machine
Competition between humans and AI is nothing new， with computers long able to challenge people in strategically complex games such as chess. The ancient Chinese board game Go（weiqi）， however， had always been considered a much greater trial for machines to overcome， with more potential board configurations than there are atoms in the universe， at least until Google took up the challenge in 2014. The company’s AlphaGo research project was created to test how well a neural network using deep learning could compete at Go.
Beating its first human opponent in 2015， AlphaGo relied on mechanisms similar to the human brain to improve its playing style from experience， before beating a world champion over five highly publicized rounds in Seoul in 2016. However， according to experts， games such as Go and chess rely more heavily on memory and computing power， which computers have in abundance， while language-based tests of the kind evaluated by SQuAD are far more difficult for AI to contend with.
SQuAD， the system used for the evaluation of Alibaba’s technology， was created by researchers at Stanford University as a means of testing the ability of machine-learning models developed by corporations， individuals and academic institutions. Launched in 2016， it consists of over 100，000 questions on more than 500 Wikipedia articles， where the answer to every question is a segment of text from a reading passage. It is significantly larger than previous reading comprehension datasets and correspondingly more challenging. According to the developers， SQuAD is unique compared with other benchmarks because its means of testing is hidden. Teams submit their code which is run on a test set that is not publicly readable， crucial to preserving the integrity of the test results. The topics of the questions are extremely broad， ranging from 16th-century religious figure Martin Luther to the immune system and from European Union law to prime numbers. One of the selected articles focuses on the Chinese Yuan Dynasty （1271-1368）， throwing out obscure questions such as“what interfered with Kublai Khan’s second invasion of Japan？” that would even have history buffs reaching for a search engine.
The competitiveness of the SQuAD leaderboard has already led to the significant advancement of novel deep learning software， as teams contend to build machine comprehension systems that can outperform those of their rivals. A number of renowned international academic institutions and global technology companies have submitted their models so far， including Microsoft， Tencent， Facebook and Tsinghua University.
Rewriting the future
While the win constitutes a major success for Alibaba and Chinese technology in general， the new systems of AI generated by the project are expected to have wider and potentially profound consequences for employment， particularly in the customer service industry.
According to Luo Si， head scientist for natural language processing at Alibaba’s research institute， machines will now be able to respond to simple questions with a high level of accuracy. “We believe the underlying technology can be gradually applied to numerous applications such as customer service， museum tutorials， and online response to medical inquiries from patients， freeing up human efforts in an unprecedented way，” he was quoted as saying on the Alizila site.
In fact， Alibaba has already been deploying similar technology in real customer service situations. When customers call Alibaba’s helpline with a question， they are usually put through to the internal computer system Ali Xiaomi. Ali Xiaomi deals with spoken and written queries， providing answers to frequently asked questions and specific inquiries pertaining to delivery status， wish lists， and recommendations. The system can automatically access account information for callers， and through the same kind of mechanisms that allowed AlphaGo to teach itself Go， Ali Xiaomi is continuously upgrading its own capabilities and learning through experience.
The technology is efficient and cost effective compared with human customer service， and， according to the company， has been especially useful in managing the huge volume of inbound inquiries during the November 11 Singles’ Day shopping festival in China. It has been so successful， in fact， that Alibaba is offering customizable， textbased versions of the technology free to all merchants on their platforms so that they can automate their own customer service. Based on this level of proliferation， observers are predicting that such technology may soon replace human-operated customer service positions worldwide. AI， the future of its application and development， was among the most talked about topics at the recently concluded World Economic Forum （WEF） annual meeting in Davos， Switzerland. Industry leaders came down on both sides of the debate， positive and negative， about the impact of AI and its associated technology on mankind. Despite his company’s recent success in the field， Jack Ma， founder and Executive Chairman of Alibaba， struck a cautious tone. “AI and big data are threats to human beings. I think AI should support human beings. Technology should always do something that enables people， not disable people，” he said at the WEF.
“The computer will always be smarter than you are； they never forget， they never get angry， but computers can never be as wise as man. AI and robots are going to kill a lot of jobs， because in the future it’ll be done by machines. Service industries offer hope， but they must be done uniquely，” Ma said.
Elsewhere at the forum observers were unanimously positive in their appraisal of AI’s potential， with many quick to point out that historically technology has created more jobs than it has destroyed. Google CEO Sundar Pichai expressed his view that “any time you work with technology， you need to learn to harness the benefits while minimizing the downsides.”
“AI is probably the most important thing humanity has ever worked on. I think of it as something more profound than electricity or fire，” he said， explaining that the technology could eliminate many of the constraints humanity now faces， helping for example to make “clean， cheap， reliable energy” a reality.
A new frontier
Since being founded in Hangzhou in 1999， Alibaba has gone on to become one of the world’s 10 largest companies with a market value of $490 billion， primarily through its successful e-commerce platforms. Toward the end of 2017， the company revealed grandiose plans to invest $15 billion in emerging technologies such as AI， quantum computing， and the Internet of Things， with its leading machine reading technology just one aspect of a much larger research initiative.
The breakthrough arrives amidst a wider climate of technological innovation in China. The World Internet Conference was held in Wuzhen， east China’s Zhejiang Province， in December 2017， where global technology leaders employed powerful rhetoric to herald a new age of consumer technology with AI at its core.
China plans to lead the vanguard of global AI research， development and application according to plans issued by the State Council last July， described by Vice Minister of Science and Technology Wang Zhigang as a major step for carrying out an innovationdriven development strategy and making the country a global leader in science and technology.
The plan stated that AI has the potential to serve as a major driver of economic growth as the shape of the Chinese economy undergoes a major transition toward innovation， capable of tangibly improving people’s lives by 2020. The release also set the goal for China to become a major global center for AI development， leading the world in AI technology by 2030.