How do we cope with this age of intelligence?

First of all, I would like to talk about the development of artificial intelligence, which in 1959, was the very first AI to be proposed, the Turing Test.

What is the Turing test?

We humans ask a series of questions to a computer in five minutes, and then the computer gives answers, and if 30% of them strike us as human answers, then he passes the Turing Test. This marked the first wave of artificial intelligence.

AI continued to be an important topic in the 1960s and 1970s, with many important algorithms coming out of that era, but then AI had its first downturn.

So what was the reason?

Because many engineers and scientists at the time realized that the algorithms and AI available at the time could only solve problems in a relatively narrow field.

So what was the problem?

It was actually that the computers were not powerful enough.

In the 1980s, there was a second wave of AI, with leading edge algorithms like artificial neural networks and expert systems. There was also a parallel increase in computing power, with the emergence of many large-scale calculators like IBM's, which allowed some real-world problems to be applied and solved.

However, AI ushered in a second wave of lows in the 1980s, when personal computers gradually began to spread, but the cost of large-scale specialized computer resources, such as the cost of building and cost is still very high. Gradually, the U.S. government began to reduce its budget and resources for this area, which led to this downturn.

Finally, the third wave of AI, as it is known, is the one we are experiencing now. Since the beginning of the 20th century, we have welcomed a series of innovations in technology such as deep learning algorithms, AlphaGo, Google DeepMind, and so on. The explosion of arithmetic and data resources has led to a dramatic improvement in algorithms.

Problems that computers couldn't solve before, such as speech recognition, image recognition, including now natural original processing and other such areas, have been very substantially improved.

?

Just now said the development of artificial intelligence wave, in fact, we can note that the improvement of algorithms and the entire intelligent technology is inseparable, but also with the arithmetic base, computing infrastructure and the amount of data, is also inseparable.

I put forward a concept, in the era of intelligence, we actually need to go through several steps: first is information technology, then big data, and finally intelligence.

We can look at the industry as we understand it to be at roughly what stage?

Let's take the healthcare industry as an example, for example, what stage do you think the healthcare industry is in?

The answer is: information technology.

Especially in China, some of the larger tertiary hospitals, such as Concordia, 301, or Peking University Hospital, are in the stage of slowly moving from informationization to big data.

One of the simplest examples is that before we went to contact an old Chinese medicine doctor, and then the director of this Chinese medicine hospital said, "We want to do big data, we have 100,000 patients' data." We were quite happy to hear that, and the amount of data is not small. Then, this old Chinese doctor took out a thick stack of medical records, he had a hundred large books, each book page is a patient's handwritten records, each book has a thousand patients, so it adds up to 100,000 patients. It was very dumbfounding, how do we analyze this?

?

Taking the medical industry as an example, what we call information technology actually includes the electronic processing of cases, hospital information systems, image management systems, etc., and at this stage, the data resources of these systems are not connected, so it has not reached the stage of big data.

If these data resources are connected, only then can we realize a patient-centered, you can see his admission records, diagnostic records, hospitalization records, but also his image data, in order to form an all-encompassing data. Moreover, in terms of timeline, such data records can form big data, and through intelligent algorithms, to help doctors to decide what kind of treatment is needed, and at what stage to do some assistance.

And this is a process that requires a continuous development for the pharmaceutical and healthcare industry.

?

For example, like KDDI recently in the hospital an application, I think it is quite good. Because I have recently been in and out of the hospital a little more, like the hospital ultrasound, need to have two doctors, a doctor to give you a diagnosis of ultrasound, and then another doctor, record the terminology he said.

Kodak Xunfei has done a thing, he is a translator, real-time record of the ultrasound diagnosis of what the doctor said, and then he automatically generate a case. This would be very effective for hospitals because it reduces a labor.

This is actually what we mean when we say, from the very beginning of the information technology, and then refer the intelligent technology to the whole business process, is a pretty good application.

Like the Internet industry, which we are familiar with, what kind of era do you think we are in now?

?

I think it is from big data to the intelligent era, so the evolution of a process in the middle.

For example, Google, he started from the web search, with a large amount of data, began to do the recommendation system, more successful, and finally launched a business model, the creation of precision advertising science. Through the behavior of our visit, browsing, to give you accurate recommendations for advertising, the next many Internet companies, in fact, are according to this path to obtain users.

Myself is doing financial big data, then for the traditional financial industry, there are actually a lot of different organizations, in different development history.

For example, we are more familiar with the commercial banks and many joint-stock banks, has now realized the construction of a big data infrastructure. ICBC, for example, has actually spent more than a decade building a bank-wide data warehouse, so it has achieved good data connectivity between its various branches and business systems. On top of that, there is a data mining platform to support its business.

?

For some small banks or financial institutions, it is still in the stage of transformation from information technology to big data. For example, a few years ago we went to do credit, in fact, what the loan examiners do, is to face a large pile of paper materials, completely relying on his experience, to determine whether to give you through. This process, not only operational risk, will also give rise to credit risk, moral hazard and so on.

In the past few years with the development of financial technology, including Internet finance, forcing the traditional financial industry to do business transformation online.

Banks are faced with Internet users, mobile Internet users, the traditional offline way has been difficult to obtain new sources of customers. From offline to online mode, on the one hand, embodied in the product innovation, the previous are mortgage car loans, the previous bank's personal business, more than ninety percent of all mortgage products.

Now there is consumer finance, such as some consumer loans and online micro-credit products, micro-finance, etc., also began to be implemented in traditional banks.

On the other hand, the previous model of relying on experts for approval has begun to shift to relying on models that, through intelligent algorithms, help to quickly identify a customer's credit needs, as well as credit risk, carrying capacity, and willingness to repay loans. Through these dimensions, a comprehensive model is formed to help with risk control.

?

In the past few years, the development of the whole financial technology is very promising for those of us who do technology, we can combine our own strengths and the traditional financial industry, and then play a greater energy.

In my opinion, pure artificial intelligence and algorithms, without the support of big data and information technology, are practically lifeless.

Back to the financial industry, we need to help financial institutions to build the bottom layer of information, and then rely on information technology big data to realize intelligent applications and algorithms.

I have recently been thinking about the business model, we are in this field of financial technology, whether it is a p2p company or bank, there are a large number of technology companies to provide services for financial institutions. The first, the mode of providing services is completely third-party form, do the technical support, including system development, data underlying construction and other services. Another model, the joint operation model, such as credit business on the injection of joint funds, or risk underwriting, to jointly operate the financial business.

?

In the future, under the premise that we have leading technology, we need to further think about how to do some in-depth combination with financial institutions to form a good business model, value output and value realization, I think this is what I need to think about, and I hope that we can brainstorm.

Follow the "Xiaobei insurance", choose insurance, not anxiety!