The end of the world as we know it: artificial intelligence and economic transformation
If you are craving a time-honored colonel Sander’s fried chicken in Shanghai, what about a whole new gen experience? In collaboration with Baidu, KFC launched in 2016 the KFC Original+ concept to offer services enabled by artificial intelligence (AI). The main attraction in my view is not the Jiangnan garden design and neo-Chinese décor themed site, but “Dumi”, the WALLE-looking Baidu’s artificial intelligence robot that uses voice and speech recognition to take your order or chat with you. You can pay for your meal with a smartphone via mobile payment services, such as Baidu or WeChat Wallet, of course. While I could not see Dumi in action when I visited the site, the experiment illustrates Baidu’s and China’s forays into artificial intelligence. In another KFC-Baidu experiment in a site in Beijing, AI technology will scan customers’ face, deduct age, gender, mood, whether the customer is a repeat business, then will analyze all the data and recommend orders based on the stream of information. Jointly, these experiments provide a glimpse into how the current Chinese environment is a fertile ground for AI experiments, into how AI is shaping modern Chinese society, and into how AI can potentially disrupt unlikely segments of global industries worldwide.
Google took an early worldwide lead in machine learning followed by other large U.S. Internet companies such as Amazon, but Chinese tech companies are trying to catch up and the experiments and projects are spreading across China. New developments in artificial intelligence are transforming every part of value chains such as production, distribution, exchange, and consumption. In the retail side of supply networks, Alibaba has reduced e-commerce customer service by about 90% with the deployment of AI-powered customer service chatbot, which can reply to millions of text queries and take thousands of phone calls from online shoppers daily. JD.com is using AI and robotics to improve efficiency in logistics and warehousing.
Alibaba is investing in the face recognition tool “Smile to Pay”, which is supported by Face ++’s technology - see video below.
Alibaba Health just unveiled an artificial intelligence service for disease diagnosis, which is geared towards minimizing the workload of mainland Chinese physicians. The country has a shortage of medical practitioners, which seems to impose a high workload on doctors and lead to a high error rate. According to the World Health Organization, there are about 1.5 physicians available for every 1,000 people in China, while there are about 2.4 in the U.S., for instance. Another AI-related medical solutions start-up is iCarbonX from Shenzhen. The company utilizes algorithms to analyze genomic, physiological, and behavioral data to produce individualized medical advice through apps. Baidu is pursuing the development of a fully autonomous vehicle by 2020 and recently announced the first version of Apollo, its self-driving car software, and its car marker partners, among which are Daimler, Ford, and 13 Chinese car manufacturers. There are experiments in all parts of the value chain and across many industries.
Image source: Molby
Experimental projects such as Amazon Go are paving the way to retail transformation in the U.S. and inspiring companies in China. For instance, TaoCafe follows Amazon Go concept, but differ by requiring a checkout step, which combines surveillance camera powered by AI facial recognition and alarm systems. Another interesting experiment is the Moby Mart in Shanghai, an autonomous mobile convenience store that integrates AI-enhanced software and mobile payment systems. The concept was developed by the Swedish start-up Wheelys Inc. in collaboration with Hefei University. You enter the store with an app, like TaoCafe, place your purchases in a smart basket or scan the items, and say thanks to HOL, the AI hologram assistant that charges your purchases to your phone. Moby runs on electricity and is equipped with solar panels. The concept is supposed to eliminate the high costs of staff and rent in a big city such as Shanghai. While not running outside of the testing grounds yet, the idea is that it will eventually use AI to navigate city streets and will be able to drive to warehouses to restock itself. Whether the concept will pan out remains to be seen, but if it does, it will certainly be an interesting addition to the daily commute - see video.
In manufacturing, AI applications represent an opportunity for productivity growth. Among the applications are the dynamic rescheduling of production operations vis-à-vis new orders entering the system, troubleshooting, and assisting material handlers in picking the correct material. It can be used to autonomously optimize operations, improve monitoring of smart networks or perform predictive maintenance of industrial facilities, and energy management. AI can potentially make inventory management systems, warehouse management systems and manufacturing execution systems faster and simpler to use. Intel for instance, deploys AI to identify common patterns among defective chips using the mass data collected from the chip manufacturing process. In China, companies such as the car maker Geely has adopted smart manufacturing, but AI does not seem to have reached traditional smaller manufacturing companies yet. The high proportion of output derived from manufacturing can potentially make AI a competitive game changer for China and other countries.
In terms of nationwide policy, somewhat similar to past U.S. policy on the technology, the Chinese government has identified AI as a critical area for advancement in its latest five-year plan. The State Council of the People’s Republic of China has released a new investment plan aimed at making China the global leader in the field by 2030. The ambitious plan involves support for a broad scope of areas ranging from manufacturing and agriculture, logistics and finance, to medicine. The plan foresees long-term breakthroughs in selected areas to support economic transformation. This plan further energizes the progress made in the area by the local AI community and makes AI another important driver of economic development and continued upgrade of the manufacturing sector. The government plan contains strong directives towards communication and coordination among universities and companies.
The importance of Chinese contribution to the field became evident recently when the Association for the Advancement of Artificial Intelligence, the world’s biggest AI research community, postponed its annual meeting by a whole month after learning that the original dates coincided with China’s Lunar New Year holiday. The reasoning is that Chinese scientists are now producing most of the research papers on deep learning. American experts continue to make most of the fundamental AI breakthroughs, but this may change as Chinese tech giants like Baidu, Alibaba and Tencent, which have access to the vast amounts of data needed for AI training through their millions of customers, inject massive investments into the technology and government support continues to build. The U.S. continues to be in the forefront at developing new ideas in fundamental AI research, while China has become good at implementing these ideas in applications. However, there has been growing investments and progress in basic research in China.
China is the world’s biggest user base and data pool for machine learning technology, which makes it an attractive ground for AI experiments. Besides government funding, venture capital funding has been a critical source of support for the sector. For instance, the CAS Institute of Automation in Beijing, a consortium of universities and firms, provides $150 million in venture capital funding to AI development. According to Venture Scanner, the U.S. is the global leader in venture capital funding of AI-related start-ups, over US$ 8 billion, while China ranks second with over US$ 1 billion. The rest of the world combined have raised about US$ 3 billion. A few weeks ago, China launched the China Artificial Intelligence Industry Innovation Alliance, its first AI alliance, to expand the industry, an initiative led by firms such as Intel China, iFlytek Co Ltd, JD.com, SAP China, and Ecovacs Robotics Co.
Artificial intelligence is behind a wide scope of advanced applications, such as facial recognition and self-driving vehicles. Many Chinese companies and the Chinese government are investing heavily and seriously on AI to obtain competitive advantages as the technology makes inroads in our daily lives. U.S. will likely experience productivity gains earlier due to first mover advantage in AI, but China may catch up quickly. Artificial intelligence was invented in the west and has been dominated by the U.S., but China is investing heavily to shape its future. While companies such as Alibaba, Amazon, Google, IBM, Siemens, among others are using artificial intelligence to transform industries in search of a competitive edge, on a national level China is investing in AI to transform its entire economy. In the process, there will be many opportunities for cooperation and competition among firms and nations, and many serious ethical and economic considerations that will need to be discussed by research, practice, and governments. For instance, as AI becomes more pervasive, how will we protect data from the increasing risk of cyber security threats and what kinds of regulations may be needed? Developments, such as the recent publication of the book of poems The Sunlight that Lost the Glass Window, produced by Xiaoice, a natural-language chat bot developed by Microsoft, raises the question about how to handle intellectual property eventually created by an AI system. This is a very powerful technology that brings many opportunities, but that also raise many important questions.