Peter Ma founded Ping An Insurance, China’s first private joint stock insurance company, in 1988, when he was 33 years old. The former start-up is now one of the world’s largest insurers, a leader across financial services in China and a pioneer in new technologies from fintech to cloud. As the company celebrates its 33rd anniversary, Mr. Ma shares his thoughts on digitalization.
My Thoughts on the New Digitalization
Digitalization has become a buzzword today, with concepts such as digital government, digital currency, digital health care and digital enterprises. It seems many traditional things can appear trendy and advanced if they become “digitalized”. However, despite the abundance of books and articles on digitalization, explanations of its theory and application tend to be fragmented or incomplete.
A human cognitive revolution
My view is digitalization is a fundamental project for the progress of human society and productivity in the 21st century. It is a revolution that enables leapfrog improvements in the ability of human cognition. Digitalization is not only a technological development, but also a revolution in the way we think and behave.
What makes digitalization so powerful? It comes from three major disruptive characteristics:
Computing power. Digitalization enables us to process massive amounts of unstructured data, enhancing our understanding of complex issues.
Cognitive ability. Artificial intelligence (AI) and machine learning mean machines have surpassed the cognitive abilities of humans in a variety of areas. They have replaced humans in key areas of analysis and practice. For instance, autonomous cars are in development that are capable of driving and monitoring the driving environment with no human driver to steer the wheel. Here at Ping An, we have already developed a powerful natural language recognition model that can interpret human speech and text. The model registered a record high accuracy of 90.4 in a global competition to answer 150,000 questions after reading 15,000 articles, overtaking the average score level of human beings of 86.8. We already use this framework for smart customer services, telemarketing, training and interviews.
Big data. The breadth and depth of the data available to us are growing exponentially, thanks to big data technology. We can access not just quantitative data, but also non-structured data such as images, videos and texts. We have an unprecedented store of information at our disposal.
The value of digitalization to society
Digitalization means we can analyze human activity according to consistent frameworks that were not available to us before. To know the truth, it is often said, one must first learn from experience.
Discovering complex associations
‘Things have their root and their branches. Affairs have their end and their beginning.’ (物有本末，事有终始)1.
New technologies also allow us to discover unknown and complex associations in human developments. Through the analysis of knowledge graphs and complex neural networks, digital technology has changed our understanding of causality and the relationships between phenomena. It enables us to go beyond isolated chains of cause and effect to detect multi-dimensional networks of association. Of course, this does not mean that the conventional single causal logic no longer exists. Instead, digitalization enables us to perceive a new type of causality: one cause generates multiple results, multiple causes generate one result, multiple causes link to multiple results or even an infinite number of causes and results. Multiple causes with multiple results is not unique to the digital age, but digitalization provides humans with the conditions and approaches to understand them.
That has profound implications for the accuracy of causal modelling in medical diagnosis, among other fields. For example, the traditional diagnosis of diabetes took into account several pathogenic factors. Yet Ping An’s technology can locate over 50 health risk factors, over 300 related diseases or symptoms, and more than 2,000 relevant drugs through knowledge graph building. We believe this will help us to identify high-risk populations for early prevention. It will not be long before this kind of capability develops the identification and management models for high-risk populations for not only diabetes but other chronic conditions, such as hypertension, dementia and hyperuricemia. By enabling early identification and intervention, we may lower the probability of incidence.
1 From Daxue or Great Learning, one of the Four Books in Confucianism. The Great Learning consists of text attributed to the teachings of Confucius, although it was written after his death. The English translation is by James Legge.
Identifying the rules of the objective world and human activities
"Although the sky is so high and the stars so distant, if we seek out [their] original condition, we can calculate a solstice a thousand years ahead without rising from our seats." -- Mencius2
The ancients recognized the existence of rules or laws governing the natural world and human behavior. No matter the size of the problem, if you can find the underlying reasons, you can predict the outcome. Through digitization, we have new tools, such as machine learning technology to calculate the development rules and future trends of the objective world and human activities.
2 Mencius (372–289 BC or 385–303 or 302 BC) was a Chinese Confucian philosopher, part of Confucius's fourth generation of disciples.
Applying new insights to decision making
“Study the underlying principle to acquire knowledge and apply knowledge to practice” – Wang Yangming (格物致知，知行合一)3
The philosophy of materialist dialectics is based on the idea that knowledge comes from practice, and in return it can guide and serve practice. The process of forming cognition or understanding does not happen at once: it is an iterative process of repeated and infinite deepening. However, human beings cannot fully understand the complex world. We are limited by our life span, brain power and memory. It is inevitable that personal experience has a lopsided, distorting influence on what we do.
Through digitization, however, humans can make more scientific decisions with the assistance of machine learning technology. Take AskBob, Ping An’s AI-driven smart doctor assistant, for example. It does better than specialists in China’s leading hospitals in terms of the accuracy and speed with which it can diagnose certain illnesses. In future, as it covers a wider spectrum of conditions, it will be promoted to remote areas and countries that have a shortage of healthcare resources and medicines, and we hope it will greatly improve public health and increase the average life expectancy in many communities.
3 Wang Yangming (1472-1529) was a Chinese statesman and Neo-Confucian philosopher.
The value of digitalization to business
Smart data analysis can help a business see what is essential in a complicated environment -- and make the right decisions in advance. Digitalization enables Ping An to compile our annual report in 34 days, giving us a head start in developing our strategy. Ping An’s annual report involves dozens of professional companies and is very complex. Off-the-shelf data governance and 100% automation of data transmission, analysis and calculation allowed us to understand the company's whole-year operations earlier, and save valuable time to formulate new strategies and take action earlier.
Improving internal management and operational efficiency
Indeed, technology helps businesses optimize all kind of processes, replacing or reducing manual work through automation – and reducing costs. Ping An’s rapid auto insurance claim solution provides a good example: customers can make claims on a mobile app within three minutes by taking photos of their cars. This fully digital, automated process is made possible by AI that has learned images of 60,000 types of vehicles and 24 million parts and components.
Risk, too, can be managed more effectively through the early analysis of financial, business and behavioral data, supported by precision monitoring and AI. We have used cognitive AI technology at our consumer finance arm, Ping An Puhui to help us assess customer credit risks and reduce credit losses by 33%.
Optimizing the customer service experience
The financial sector is well suited to use digitalization to improve services and customer experience. AI also helps us at Ping An to achieve a more personal relationship with our more than 220 million customers. Human labor alone cannot deliver our service standards to such a vast customer base. We call it “heartwarming” finance because we are disrupting traditional financial services delivered in a cold-hearted and infrequent manner to provide our customers with services such as healthcare. We rely on digital healthtech capabilities to offer online and offline healthcare services with our “AI + Human Doctor” initiative and provide our customers with care and services 24 hours a day, seven days a week.
Replacing repetitive, labor-intensive work with machines
AI can manage simple, repetitive tasks with efficiency and accuracy, reducing operation costs. AI service representatives are used across the Group: in 2020, the amount of services provided by AI service representatives grew 41% year-on-year to 1.93 billion times, representing 82% of Ping An’s total customer service workloads.
Digitalization will continue to disrupt all areas of life in the 21st century. Sitting in my office, I’m always worried about falling behind and becoming obsolete in the digital age, when the market competition will only grow more brutal and winners will take all. We have to remain vigilant and alert to the opportunities and challenges that lie ahead to continue our sustainable development.