Can you tell us more about MioTech. What is it? What do you guys do?

MioTech is a platform that is committed to bringing data intelligence by using cutting-edge artificial intelligence to financial institutions. So I work both in finance and in technology. I see both the richness of financial data and the cutting-edge technology that's available in the world. So when I came back to Hong Kong in 2016, I decided to found MioTech so that it enables financial institutions to embrace data and to unleash their team potential.

In the financial sector, we already have some pretty advanced tools that we're using to crunch data and to help us do research. What are the weaknesses you see in these tools that you want you to fix?

I actually see innovation in three different phases. The first phase is what I call innovation 1.0, it’s digitization, where you bring offline stuff, online, you connect people, or you bring offline processes online. The second wave is what I call big data. It's a relatively short-lived era because people start to realize once you bring all these offline stuff online, then there is a lot of data that you need to analyze. And Big Data is simply by putting all the data together, what they call business intelligence. And right now, we're at the third era. It's called artificial intelligence. Because data is overwhelming, the traditional ways to analyze data is no longer very effective. That's why we need artificial intelligence.

So how is your product different? How exactly can it help solve those problems that you just mentioned?

Think about a digitized product where it only brings data online. There's no way for you to interpret it. But for AMI, we spent the past two years building its graph engine, building its natural language processing system and building a topology based graph mining tool so that we know the interconnectedness of all different information nodes and then we can mine the data and get the true intelligence. For example, if you ask AMI a question like what is trending in say electric vehicle. At a click of a button, AMI would tell you all the important information you need to know and possibly some personnel or companies that you need to pay close attention to.

So from the sound of it, is AMI the SIRI in the financial sector?

I think Siri or Google or some of the other search engines has its flaws because the financial world is more interdependent. For Siri, when you ask Siri a question, it returns one specific piece of information. When you ask Google, it returns a list. Whereas if you ask AMI, it actually maps out the whole financial world and to help you highlight the most important piece of information that you should pay attention to.

AMI is relatively new. Can you tell us more about it? How does it work and can we really use this technology to measure innovation?

I think of innovation as the future delivered right now. So AMI helps you to take a sneak peek at it. So how AMI does it is rather complicated. But let me try to explain it. Think of AMI as a two-year-old right now. Or let me give an example of Alphago. Think about Alphago when it was first launched years ago. Engineers saids that it has a two-year-old equivalence of intelligence and as it plays against more and more top players, it gets smarter and smarter. That's the essence or the art of AI. Compared to traditional rule based engines, we incorporate a lot of, for example, deep learning algorithm, like convolutional neural network. Put it in simple words, it actually learns from both our customer behaviors and from the information that we feed to AMI.

How objective is it, for example, what does it mean to investors if AMI says companies like Apple is more innovative than Samsung?

It actually gives you a perspective of what AI thinks about things. The inner structure, the inner construct is like a human brain where it defines neurons, the information passes through different neurons. So what does it tell? It tells you that based on our algorithm or technology, Apple sounds more innovative as compared to Samsung. And let me remind you that every month we get 1.5 million online narratives. We have in our system we have 120 million company profiles (Update: new data injection of 175 million) and 5 million personal profiles. So our result is based the synthesis of all these kind of information that a human brain is not able to process but a machine can.

I would like to bring the discussion back to artificial intelligence. I would like to get your perspectives on what value do you think AI could bring to economies, societies and businesses.

For example at MioTech, we want to make sure that AI is not only a disrupter but also an enabler. More importantly I would say an enabler. It enables human beings to be more efficient and to make better decisions and reduces the cost. A lot of people, for example, worry about the ethics of AI. But at MioTech we take what we call a human centered AI approach. I actually personally learned it from my home school Stanford. So they have a set up a human centered AI principles. One of them, for example, is that it needs to benefit the overall society. So when we design AMI, for example, in the financial information industry there are people who are working on technologies that can help speculators. But in our mind, the biggest value of Finance is to allocate resources efficiently to the companies or to the sectors that needs the financial resource. So we took a very value investing approach, where a lot of the insights generated from AMI, we want to help people to find out values in the supply chain, in the industry, in the sector or any of the emerging leaders that they traditionally haven’t be able to see.

Where do you think in Asia, AI is going to make the biggest impact? Can you give us some examples. How are your clients benefiting from the technology, of AI or AMI?

In our mind, it is the finance industry. First of all, in terms of data quality and the amount of data, the financial industry has the largest amount of both structured and unstructured data available. That's a very sweet thing in artificial intelligence. Basically you can train the brain with much larger data set. if your two-year-old reads more books, he would become smarter compared to other kids. So there's a lot of information in the financial industry that is available for us to mine, for us to teach our artificial intelligence engine. So I think that's where the first batch of applications are going be developed.

For AI to grow, we can't just develop AI. There are so many other things we need to think about, for example cyber security. Where are we at in terms of cyber security in Hong Kong? Is it at the level where it can support the growth of AI technology yet?

I was having a very interesting conversation with a friend of mine and let me share with you this this example. So we were arguing about an analogy. We think that constructing a building is similar to building a technology product. However, my friend argues that when you construct a building, after you construct it, you would put all the security measures. So that's not equivalent to security that we were talking about. To me, security is the foundation of the building. If the foundation is weak, you spend years and years building an awesome product and all of a sudden you figure out that there are security leaks and the building just collapses. So to us, especially to MioTech or to a lot of my fellow entrepreneurs who want to build a great product, we spend more time on cybersecurity than anyone else. We’ve spent the past two years trying to develop AMI and the first year, we spent almost half amount of our time building the security measures including deployment, in the encryption, internal compliance. So that we can make sure that it doesn't go wrong and the building won't collapse. And then the rest of the time is spent enhancing our technology so that we can now say we have a great product.