AI in China: Key players, focus areas and funding Business Finland China
Executive summary China aims to build a 1 trillion yuan ( 126.4 billion) AI core business by 2030, which will potentially stimulate as much as 10 trillion yuan ( 1.26 trillion) in related business. Investors poured $4.5 billion into more than 200 Chinese A.I. companies between 2012 and 2017 and half of the AI investments world wide went to China in 2017 In November 2017, the Chinese government announced that it plans to build four national AI open innovation platforms by relying respectively on Alibaba s smart city technologies, Baidu s self-driving technologies, Tencent s AI-enabled medical treatments, and iflytek s voice-recognition technologies. On education, health, transportation and manufacturing, Chinese players have being moving to create their own ecosystems and platforms for AI application. At the current stage, AI+education appears more fragmented than the other three fields. Successful AI start-ups from China have secured record-setting levels of financing, backed by both government and corporate actors. Government-backed/managed funds focusing on AI are being established on both national and local levels as well. Despite the policy and investment support, presently China still needs to catch up with top AI countries on its capacities in hardware, research and commercialization. 22.8.2018 2
The BAT battle
BAT (Baidu-Alibaba-Tencent) led national team has been tasked with building 4 open innovation platforms - AI assistant/platform DuerOS - Self-driving platform Apollo - Deep learning platform Paddle Paddle - Platform of Artificial Intelligence (PAI 2.0) - Tmall Genie voice assistant - Customer service chatbot Dian Xiaomi - Tencent-developed cloud service - Open-source computing platform Angel - WeChat AI - Robot journalist Dreamwriter - iflytek Open Platform 22.8.2018 4
Regulations loosening for autonomous vehicle testing - Baidu currently leading but other big players follow closely China s Ministry of Industry and Information Technology (MIIT) has released national guidelines for smart internet connected cars, which allows local authorities to arrange road tests based on their specific conditions and coordinate all parties more conveniently. Baidu has been offering its autonomous driving software platform, Apollo, to companies for free. Baidu hopes to collect data from usage of the Apollo platform by its partners, including car makers like Ford, car-component makers like Continental, and chip makers like Nvidia. In addition to Baidu, both Alibaba and Tencent have been conducting tests of self-driving cars. In April 2018, reports said that Alibaba had been running road tests of autonomous cars on a regular basis, and was looking to hire 50 more self-driving specialists for its AI research lab. In May 2018, it was reported that Tencent obtained a license plate for its self-driving car from the Shenzhen Transport Bureau, which gives it permission to conduct relevant testing in the streets of Shenzhen. 22.8.2018 5
Alibaba s data-hungry AI is controlling (and watching) cities in China Alibaba have announced that they will invest 15B USD (until 2020) in seven research labs (Beijing, Hangzhou, San Mateo, Bellevue, Moscow, Tel Aviv and Singapore) that will focus on AI, machine learning, network security, natural language processing and more. City Brain monitors every vehicle in Hangzhou, and Alibaba has reported that it have helped reduce traffic jams by 15 per cent City Brain works by letting artificial intelligence (AI) control a city. Large amounts of data are gathered, processed by algorithms in supercomputers, then feed it back into systems around the city The goal is to make decisions in areas such as road planning, bus routes and frequency, and the length of time a particular red traffic light should be on, to increase traffic efficiency Macau and Malayisia s Kuala Lumpur as next implementations in the pipeline 22.8.2018 6
Tencent aggressively investing in AI for healthcare, teaming up with start-ups with expertise It is estimated that Tencent has invested 2.57 billion into 33 health-related start-ups at home and abroad since 2014. In China, the company is also opening clinics and an artificial intelligence-powered laboratory. We Doctor, the Tencent-backed Chinese online healthcare services firm, has completed internal business integration and scaled up its four verticals medical cloud, healthcare, medicine, and insurance. The company has raised 428 million pre-ipo investment, the biggest for health tech companies in China. This puts the valuation for We Doctor at 4.71 billion. UK start-up Babylon is partnering with Chinese tech giant Tencent in an attempt to take on the Chinese market. Babylon has an AI healthcare mobile app that allows patients to self-assess their symptoms prior to contacting a physician. Now it is teaming up with Tencent to deploy its AI app on WeChat. 22.8.2018 7
AI Development in other areas
AI is used to deliver more effective teaching to the masses, and speed up evaluation Chinese tech firms have been using AI to drastically improve the efficiency of education content delivery and diagnosis of students response to test questions. Liulishuo, an AI-enabled English learning app, has launched their AI English teacher, who has been proven to triple the learning efficiency compared to human teachers by CEFR (Common European Framework of Reference for Language). Liulishuo now has over 70 million users, covering 379 cities in China and 175 countries around the world. Master Learner, an online education start-up, has been developing an AI-enabled platform to review students reponses to test questions. This AI-powered system was trained using a large numbers of math, physics and chemistry questions asked by Chinese middle school students on a daily basis. 22.8.2018 9
Face/voice recognition is deployed to collect data used to improve the learning process Hujiang, one of the largest online education sites in China, is working to utilize image recognition and voice recognition to capture students facial expression and feedback to improve the interaction between pupils and teachers in online classes. TAL Education, a NYSE-listed company offering tutoring services, is working with Stanford University in the US to further advance the use of AI in its online tutoring programme. TAL and Stanford formed an alliance in early 2018 to explore applications of AI in assessing students comprehension during the learning process. TAL s online platform has been capturing data on students level of understanding by using facial and voice recognition technologies. Its team can use the data to give tuition support after class, or deliver more personalised education content to users. 22.8.2018 10
Haier cultivating an AI-enabled ecosystem for intelligent manufacturing Haier s COSMOPlat industrial internet platform creates an ecosystem that allows customers to create personalized orders for products by Haier In 2017, COSMOPlat became the world s largest mass customized solution platform with total transaction reaching 313.3 billion yuan (41.79 billion EUR) and 4.12 million orders. In February 2018, this platform became China s first national-class industrial internet platform approved by the National Development and Reform Commission (NDRC). 22.8.2018 11
AI Funding, Research and Conclusions
Funds focusing on AI have been established in China on different levels National The Fund for Industrial Restructuring and Upgrading is a new central budgetary scheme started in 2016 and managed by MIIT and MOF, with the goal of supporting industrial restructuring and technology upgrades in line with the Made in China 2025 roadmap for manufacturing. While the full scale of this support is unknown, in 2016, this budgetary scheme allocated 2.783 billion RMB (370 million EUR) to projects in smart manufacturing standardization and new model applications alone. This fund did not exist prior to 2016. The Chinese National Venture Capital Fund is a 200 billion yuan (26.68 billion EUR) state-controlled fund set up in 2016 to invest in China s start-ups. It led the high profile financing of AI firm Face++. Regional Shanghai In November 2017, it was reported that the local government in Shanghai unveiled guidelines to develop the AI sector by granting special funds and enhancing support. The Shanghai local government has not yet disclosed how much funding will be granted to the industry. Hangzhou Sinovation Ventures announced in November 2017 that the company will establish a 4 billion RMB (530 million EUR) fund in Hangzhou, focusing on AI. As of April 2018, a RMB 10 billion (1.33billion EUR) government-backed blockchain fund called Xiong An Global Blockchain Innovation Fund, has been launched in Hangzhou to invest in innovative firms using blockchain technology. Others In December 2017, China-based accelerator firm COMB+ announced that it was planning to launch a 65 million fund aimed at helping promising international AI firms enter the China market. The fund was reportedly backed by government funds, government institutions, private enterprises and big corporates most of which are from China. The fund is looking for startups with a proven business model in local market who would like to go to the Chinese market, and can help them further develop and upgrade their tech so that they can apply it to the Chinese market. For deals, COMB+ will be looking to provide 1-2 million typically, with follow-on capital available. Individual investments will not top 5 million per deal, according to the CEO. 22.8.2018 13
Emerging AI unicorns work in voice/face recognition, robotics, education and media Name Sector Financing Total (USD) Valuation (USD) Key Investors Toutiao 今日头条 news $3,150 M $35B General Atlantic, Sequoia Capital, CCB International, Beijing HQ Location Ubtech 优必选 robotics $942 M $5B Tencent, CDH Investments Shenzhen SenseTime 商汤科技 face recognition $720 M $4.5B Alibaba Beijing Face++ 旷视科技 face recognition $460 M $2B China State-Owned Capital Venture Capital Fund, Ant Financial, Foxconn Beijing Mobvoi 出门问问 voice recognition $255 M unknown Volkswagen, Google, Sequoia Capital Beijing icarbonx 碳云智能 healthcare $255 M $1B Tencent, Zhongyuan Union Stem Cell Bioengineering Shenzhen Unisound 云知声 voice recognition $230 M $1B CLP Health Fund, 360 Beijing Cloudminds 达闼科技 robotics $130 M unknown Softbank Group, Foxconn Beijing Roobo 北京智能管家科技 robotics $100 M unknown iflytek Beijing Liulishuo 英语流利说 education $100 M unknown CMC, Wu Capital Shanghai Appier 沛星互动科技 marketing $82.5 M unknown Softbank Group, Line Group, Naver Group Taipei Aispeech 思必驰 voice recognition $45 M unknown undisclosed Beijing 22.8.2018 14
China has unparalleled access to data for its AI development, but is still lagging in other areas An Oxford study showed that China trails the US in every driver of AI development (hardware, research and algorithm, and commercialization) expect for access to data China s AI research capacity still suffers from a lack of experienced AI researchers and fundamental innovation in algorithm development Processor and chip development may be the biggest challenge for China s AI plan, due to high initial costs and long creation cycle BATs are collaborating with Chinese universities to boost joint research capacity and domestic talent pool In February 2018, iflytek established an AI school at the Chongqing University of Posts and Telecommunications In April 2018, Alibaba and Tsinghua University launched a joint lab focusing on human-computer interaction technologies In April 2018, Tencent Cloud and Shangdong University of Science and Technology jointly established a school for AI teaching and research at the University 22.8.2018 15
Conclusions China wishes to become one of the leading global hubs of AI development, with the help of strong policy support and significant investment in the area China s tech giants (especially BAT-companies) are given a national task and are investing in AI platforms in autonomous driving, healthcare, smart cities, intelligent manufacturing and voice recognition & intelligence (often applied for education), which opens up potential cooperation opportunities for Finnish companies in these areas. China is currently trying to catch up on its capacity to conduct research and innovation and supply core hardware for AI development, and the gap (to US) is closing as more investments are poured into these areas 22.8.2018 16
THANK YOU Arto Mustikkaniemi Head of Innovation, Business Finland China arto.mustikkaniemi@businessfinland.fi Luke Li Advisor, Business Finland China luke.li@businessfinland.fi