AI: Already all around us, and touching many aspects of business and everyday life

So called ‘robo-advisers’ – asset management websites or apps that offer financial guidance and investment advice based on algorithms – are fast becoming a popular option

PUBLISHED : Friday, 16 December, 2016, 4:31pm
UPDATED : Friday, 16 December, 2016, 9:22pm

For many people, the expression “artificial intelligence” might still seem like scary science fiction, conjuring up images of a world run by faceless machines and robots.

Experts still say it is likely to be some time before it replaces you or your colleagues, but the fact is that AI in its various forms is already all around us and penetrating many aspects of business and everyday life.

Let’s start with your (days mainly digital, these days) wallet.

The yuan’s recent depreciation has prompted more people to search for investment products that offer better yields. But instead of going to a bank, queuing for hours and being offered simple advice from a salesperson, many have been turning to their mobile devices to search for more convenient alternatives online.

So called “robo-advisers”, a label for digitised asset management websites or apps that offer financial guidance and investment services based on algorithms, are fast becoming a popular option.

While not exactly talking to a robot, users are advised based on a simple questionnaire.

Tap in your age, background, assets and risk tolerance, and algorithms deliver options that might best suit, explains Zheng Yudong, chief executive of Xuanji, the robo-advisory unit of Chinese financial technology platform Pintec Group.

“By answering some hypothetical questions, AI is able to work out your risk tolerance,” Zheng said.

“An example question might be ‘how much loss do you think will cause you insomnia?’ The more details you provide, the more effective and accurate the algorithms can be.”

From a portfolio generated by the robo-adviser, users can pick different products that track stock indices, currencies and various other types of investment that might hopefully better manage their individual circumstances and earning targets, he says.

“The demand for pure human financial advisers continues to shrink as their charges are high and they are only able to service a limited number of clients – not a problem for AI-powered robo-advisers,” Zheng said.

He said continued stock market volatility, rising property prices and mounting concerns over China’s economic slowdown were already prompting rising numbers of mainly middle-earning investors to look into more dynamic investment portfolios than just betting on a single stock.

The demand for pure human financial advisers continues to shrink as their charges are high and they are only able to service a limited number of clients – not a problem for AI-powered robo-advisers
Zheng Yudong, chief executive, Xuanji

“China’s growing middle classes have largely been overlooked by human advisers at banks and wealth management firms as they don’t invest large enough amounts for the advisers in fee terms,” Zheng said. “But they still need financial advice and many are finding robo-advisers the answer.”

Already, robo-advisory services have funds under management estimated at about US$30 billion globally, but Zheng predicts that figure could mushroom to US$3 trillion to US$4 trillion by 2020, which would represent about 5 per cent of the world’s wealth management market.

Their use is already well-established in mature markets such as the US and Europe with investment banks and financial institutions flocking to invest in the technology, but the sector is still in its infancy in China.

Although robo-advisers are more convenient and cheaper, Zheng adds their actual range of advice is limited and they are unlikely to ever lure high-net-worth individuals away from their human wealth managers.

“Many complicated issues will always require the personal touch, such as family succession issues or the likes – considerations that a machine or algorithm will never be able to handle,” he said.

Their real advantage, however, comes in being able to process huge numbers of customers looking for relatively basic products and guidance, and they are already boosting the wealth management sector overall.

Hyde Chen, an analyst with the chief investment office of Swiss bank UBS – itself a large user of robo-advisers – insists that AI technology in financial services will never fully replace human advisers, but their ability to handle volumes of simple, repetitive inquiries more efficiently leaves actual advisers more time to concentrate on more complicated demands.

“Jobs with three characteristics are at the highest risk of being replaced: low-skilled roles, repetitive tasks and jobs that are predictable. AI [applications] will be doing these jobs more effectively in future,” Chen said.

UBS estimates that in the next few years globally up to 75 million jobs could be affected or eventually replaced by AI, specifically in areas such as sales, marketing and customer services, but also in more unlikely sectors such as stenography and even journalism.

Social media content aggregator Toutiao, which literally means “today’s headlines” in Chinese, is effectively an online news outlet that already depends largely on AI technology.

Accessed through a mobile app, customers receive their news feeds and alerts based on algorithms (in a similar way to robo-advisers) based on basic information such as age, gender, location and which smartphone model they use.

Toutiao’s AI systems recommend customised content and alerts for users – a very small number of items are actually written by robots.

A spokesman for Toutiao’s technology team uses the examples of a smartphone user based in Beijing’s Wudaokou, in the city’s northwest district of Haidian that is close to a number of universities and research institutes, and one in Shanghai’s Lujiazui financial district.

The former might receive news on Tsinghua University, the education sector, or research papers while the latter could be recommended news feeds about business and finance.

“The more information we can receive on users’ most-read content, the better our AI can make accurately targeted notifications, driven by algorithms and other forms of so-called ‘deep learning’,” the spokesman said.

Toutiao already claims to have more than 66 million active users, who read an average of three billion articles per day. Its content comes from a number of registered accounts, from social-media-based news feeds or blogs created by individuals to media outlets including China Central Television news and Xinhua, the government-run television station and news agency.

Humans are indispensable when detecting fake news, but they work well together with our AI technology
Spokesman for Toutiao’s technology team

It also delivers news feeds from other government agencies such as Beijing’s information office to more specialist offices including the National Copyright Administration.

But much like the wealth management industry’s AI-based services, its great advantage comes in being able to handle huge amounts of information effectively and repetitively.

Toutiao also claims its technology is able to filter possible fake news although it needs the human touch for that. A lot of the false news is identified through alerts back from users of the system.

“Humans are indispensable when detecting fake news, but they work well together with our AI technology,” the spokesman said.

While mobile apps such as are shaping people’s reading habits, AI is also changing what they write and how they use their voice.

Voice recognition has become a must-have function for many smartphones, and now users can also talk to their own devices, which can then write out what they say into messages.

Chinese internet giants Baidu and Alibaba, which owns the South China Morning Post, have been investing big in AI technology, and voice recognition is fast becoming one of their main focuses.

Baidu first began developing its AI system in 2013, the “Baidu Brain”, which is now powering voice recognition as a significant function in its Baidu map.

Alibaba has also expanded its AI application, “ET”, into areas such as traffic control and real-time voice recognition functions that assist in customer service support for e-commerce users.

Alibaba’s cloud computing unit showcased the company’s latest speech recognition feature at an industry conference in October, by using it to subtitle speakers’ remarks on a huge screen in real time.

With the development of voice and image recognition and other applications, AI will revolutionise many industries, from manufacturing to agriculture, to medicine
Zhou Jingren, chief scientist, Alibaba’s cloud computing unit

The unit’s chief scientist Zhou Jingren claims the accuracy was 97 to 98 per cent. As yet, it is limited to Putonghua but more languages and accent recognition will be built in very soon, including Cantonese. It also has a lot to learn, he adds, on recognising synonyms and voice tones that may change the meaning of some words.

“With the development of voice and image recognition and other applications, AI will revolutionise many industries, from manufacturing to agriculture, to medicine,” he said.

Lee Kai-fu, a veteran technology investor who founded venture capital firm Sinovation Ventures, is convinced that the use of AI will change the wealth management, banking and insurance sectors forever, and already the data that has been gathered by the systems in use has created a solid foundation for rapid growth in future.

He told a recent industry forum, organised by start-up service platform, that China was in a prime position of having fewer constraints compared with Western countries when it came to privacy and being able to collect and use personal data, making it easier for companies to expand and develop their AI applications.

“The technology will bring significant changes to many sectors,” he said. “AI-powered voice and face recognition applications will replace many different types of jobs, from security guards to stenographers, in the coming decades.”