Panel Discussion II: Adopting AI Technology for Enterprise
The second panel discussion of the day is attended by Anthony Arundell, Director, Sustainability, Smart City, Frasers Property Holdings Thailand, Chang Foo, Managing Director, Tencent Thailand, Tortakool Khongsap, Ph.D., Deputy Managing Director of Thai Beverage Logistics, and Prinn Panitchpakdi, Country Head, CLSA Securities, moderated by Michael Araneta, Associate Vice President of IDC Financial Insights Asia Pacific.
What are the use cases of AI in your organizations?
Starting with Anthony Arundell, Director, Sustainability, Smart City, Frasers Property Holdings Thailand, “Frasers is developing one of the largest projects in Thailand; a city within a city called One Bangkok, consisting of hotel, residential and retail facilities. One of the key challenges is that this is a green field development, so we want to do something very exciting, from smart city, smart living to sustainable principles. When we looked at some of the technologies we need to deploy, with the large amount of cameras, and video feeds, the question is how do we monitor those effectively and efficiently to give one of the best and most appropriate public safety to all the occupants of One Bangkok. So, we looked at AI, on how we could actually process video image recognition and analytics, and consider how can we deliver better services. We also look at how we can gather and analyze information from our IoT platforms, analyze it and then give that back into more real-time and predictive maintenance scenario to make sure we provide better services to our tenants.
One case, for example, is how we manage our Car Parking system, how do we manage the traffic in our Car Park, and how do we optimize our parking system to get the best parking solution for our tenants. So these are some of the challenges we looked at, which always come back to the business perspective of how we can apply technology to get the best solutions for our occupants.
A lot of the technology we are trying to deploy in our development hasn’t been used before, so we need to get the businesses to say, yes, we believe this is going to work. There are a lot of challenges in that internal business process to get to the goal. Data is essentially the key, and it revolves around the ethical management of data. And this is something that the industry needs to police ourselves as we move forward and not just to rely on regulations from the government.”
Mr. Prinn, Financial Services also has a lot of data and privacy issues; what are the opportunities for AI for Financial Services, but also for CLSA Securities in specific?
To that question, Mr. Prinn replied, “CLSA is now owned by Citic Group, one of the largest conglomerate firms in China, so you can imagine with China being at the forefront of investing in AI and using the abundance of data they have in China that this potentially would allow both the country and the businesses in China to be able to leapfrog their competitors.
We set up a Financial Technology Arm or FinTech, and with that goal, we have been investing in specific artificial intelligence projects around the region, and also looking to partner with some AI specialized funds. On the business level itself, there has been a lot of disruption in the financial sector. The commission rates in investment brokerage has been continuously dropping worldwide, but AI enable us to enter new areas of products and services. There have been foreign brokers giving services on private funds in the form of a robo-advisor, where behind the scenes of the robo-advising, they are collecting data and recognizing patterns on past investment strategies to inform you of the various strategies they can take. These strategies can also be event specific, or behavior specific, depending on your patterns and activities on the platform, or those of others. And while this data may not be perfect, they are able to extract some patterns of the emotions of traders in response to various events.
Employment-wise, AI has also brought upon dramatic change, where before we might need 100 traders to execute the clients’ orders, today, we only need 3 or 4 people to carry out those same tasks. So artificial intelligence allows us to make use of the data in a more effective manner, and in a more timely fashion.
Another case I took in my office the other day is on Social Impact Investment, where you could input the criteria of the corporate governance score, the managerial quality score, accounting standard score, transparency, disclosure, added in sustainable angle, such as the company’s environmental impact policy, gender equality, and even score how the corporate is helping society at large. After which a ranking is assigned in a very efficient manner through AI technology, on the level that transactions can happen so quickly that you can not prove at a certain instant in time whether they have actually been holding the stock or not.”
So with the implications of the loss in employment of your brokers and the lack of verifiability, do you take a position that AI might in fact have a drastic impact on the financial world?
Mr. Prinn added, “For sure, it already has. You see the job employment loss, the fee structure that we had to change, the model that we operate had to change dramatically, the way we interact with our clients, and the way we work in our teams. The regulations don’t make it any easier, because the rule of laws in the financial world try to segregate each institution and function and how fund managers can trade and give orders, etc.
But in the future, I am hopeful for the future generations to be able to use the tools available to them wisely to achieve a better sense of financial planning and activities. On the other hand, these tools would also increase market volatility. For example, you can see in the last few minutes of the New York Stock Exchange where sell and buy orders are executed in a matter of seconds, that clearly those are the execution of robots, not humans, so the market is very volatile.”
So on the net, you think that AI is negative for financial services?
To which Mr. Prinn replied, “Like any new technology tools, it is a double edge sword. It brings challenges, so it’s up to the humans to learn how to use the new technology and what opportunities it can bring. For example, even in the financial industry, employment is sometimes handled by AI, where specific candidate qualifications were input into the system to screen employment applications.”
Looking for some good implications of AI from Tencent, any good things on the use of AI for Tencent?
Chang Foo, Managing Director, Tencent Thailand replied, “I just want to make a comment on the example Mr. Prinn just made based on an article I read that Citibank is about to eliminate 20,000 jobs, so it is clear that AI is quite prominent to banking, but to Tencent, we have been aggressively pushing AI for over 10 years, and you can see that in all our products. AI is also part of our DNA, for example WeChat, games, and videos. To give an example of what we do in Thailand, Joox, the number one music streaming application in Thailand, started off being a late comer in the music streaming application frontier. So how did we achieve this? In everything we did, we track data. We track everything the user does. On a daily basis, we have millions of active users spending over 90 minutes to 2 hours on the application. This allow us to not only know what songs they are listening to, but we have co-catalogue of music, and are able to identify segments and associate certain types of music to certain type of personas. So with the data we have been collecting over the past few years, we are able to provide recommendations to engage the users more on a daily basis. So this is an example of how we use data and continuously using machine learning to recommend the right songs to the right people.”
Do you think that this goes towards the discussion around privacy?
To which Mr. Foo responded, “I think there is a fine line like you said, but I do not think we are abusing data. I think that if we are advertising too much, it is getting towards your earlier examples about target and sending advertorials into your face. But for us, we are only recommending songs or content that they want to consume. The users have a choice to skip, to stop listening, or to select what they want to listen to.”
Thai Beverage Logistics, you’ve got a different set of use cases.
Dr. Tortakool Khongsap, Ph.D., Deputy Managing Director of Thai Beverage Logistics began sharing his perspective on AI use cases, “We have several projects based on AI, but I want to discuss the few projects we are working with Carnegie Mellon. One of the project, called Agile Supply Chain, relates to how we transport our products. The project actually evolved from data analytics, analyzing the transportation routes. This project has been progressing over some time and is definitely not a quick win. It is worthwhile of a Ph.D. dissertation.
Another project that will be completed soon, this year, is on a recycling plant, using image recognition to separate good and bad bottles within 0.01 seconds, which is much faster than the original methods. Anything to do with improving the speed of computing is improving productivity and efficiency on a massive scale. AI is nothing new, it is essentially the principle mathematical products.”
Mr. Michael then summarized the various representation of how AI can be applied to an organization, from optimization to improving speed, increasing the rate of personalization, achieving things you may not have been able to do before, to how decision making processes in the financial industries can be executed in a much faster level. To which Mr. Prinn added that the ability to use new technology to bridge the digital divide gap or not to create one is very important including 1. Social credit scoring 2. Peer to peer lending platform and 3. Village fund lending based in Chachoengsao.
On the subject of AI for the good of society in general, where do you think the opportunities are for Thailand, Anthony?
“For Thailand, we talked about how we could have a social scoring for people, which is a good point. Other areas that we really need to look at are public safety and security. AI could actually help a lot on that area.” Mr. Anthony replied.
To which, Mr. Foo added, “With Tencent Cloud, we actually have a city in China where we have CCTV, and with public safety, we have been able to identify criminals fairly quickly whenever they are in a public space, and that helped reduce crime significantly.”
Dr. Tortakool added, “We promote safety and sustainability, and we have been doing that with the majority of the Thai Beverage Group. With AI, we have opened up a lot of opportunities, new jobs, and new capabilities. For example, we have been using robots and AI in recycling plants help out the manual labors, and allow them to do something else.”
So in this panel, we have established that AI could be good for business, and also good or society. Will it be bad for society?
Dr. Tortakool was the first to respond, “It depends on who uses it. Say, if it steals my identity and post something bad on LinkedIn, it could be bad.”
Mr. Anthony added, “Any technology could be use for good or bad. Say, if you start using AI to identify individuals or groups and start prejudicing against those people, that would be a step in the wrong direction. As a community or a group, we need to try to enforce that socially as well as by regulations.”
Dr. Tortakool added, “If you know everything about everyone, then there is a question where the fairness would be. Take insurance, for example, when you pay an insurance premium, you place your money into the pool assuming that the risk is the same. But if you know everything about everyone, then the questions arise whether or not it is fair for me to pay a certain amount.”
Mr. Foo then responded, “I think it is inevitable, we just need to accept the fact. However, if there is more good than bad, then I think it is okay, since it will help all industries, especially healthcare. For example, AI can help with patients diagnostics following a certain protocol before they get to a terminal stage, and so if it helps over all, we just need to accept the fact that it is here.”
Lastly, Mr. Prinn added, “We have discussed about how it definitely brings more efficiency, eroding some of the repetitive tasks that humans do, doing them at a faster speed. In healthcare, for example, while AI may help you get through the X-Ray results faster, you still need the human. The whole governance issue is being called into question. This is essentially a challenge for the Government to regulate. For example, we are starting to hear discussions about universal basic income, where in the future, we will begin to see new types of employees that are not human. For example, today, you are trading against robots in the stock market. Like it or not, a lot of jobs will be lost, especially for those people who refuse to be reskilled, retrained or change their mindset. This is a challenge for the Government on how you would be coping with the labor issues.”
What are the things that Thai Enterprise should do now in 2019 that would yield some momentum for AI?
Calling on to Dr. Totrakool to be the first to respond, “You really need to have a business strategy, say, if you want to do something transformative, it is not going to happen within a year. For Thai Beverage, we are already planning for 2025, so we look at the long term and we expect that within a few years, we will start to see some results. To make sure that you will be competitive for the years to come, you really need to craft a long term strategy. By 2025 or 2030, most companies will eventually have to be a technology based company. This is unavoidable.”
Mr. Foo then added, “I think that a lot of organizations want to implement AI, but they don’t really understand AI. They need to understand the practicality of what is capable through AI. I think first and foremost is if there is an organizational will to implement AI, you need to understand what AI is, and then build the team around that, get the resources, but first, you need to start with something that you understand. Address it with a problem statement, put the resources around it, and then do a pilot project.”
Dr. Totrakool then added, “It’s the business strategy first, AI strategy second. Otherwise, you would not know what to do.”
Mr. Anthony responded, “Very similar, I think you would need to define your use case, somewhere you can start and get tangible results. Also, you need to prepare your organization for the change. Other issues around technology and software is that there are a lot of software from Microsoft, SAP and Tencent that you can actually use to accelerate AI adoption into your business. But the key to success is then how your businesses could take that technology and then adopt it.”
Lastly, Mr. Prinn added, “For Thai companies, I strongly urge you to look beyond from your network and Thailand. Because obviously Thailand is not a leading expert in AI. You need to look at your pain point and then consider how AI and Blockchain can help, because AI is not going to be a silver bullet that will cure everything. Thailand is not yet at 4.0, but we are at 0.4. By that I mean that in Thailand, we do have emotional intelligence, we do have empathy towards our colleagues. Use this to leverage your business, your platform, your technology. I will give you an example of a company called LocalAlike, a local tourism company started from the root people in the community and that is beginning to use AI to target a specific tour package towards a specific group based on what they like. They know the price point that relates to the customer group. Think about how you can leverage off that, because Thai companies have many strengths already in these aspect areas.”
Where should they start for typical Thai company?
Mr. Prinn replied, “China, Russia, Israel, London, you really need to look outside, but start somewhere near like China, for example.”
Dr. Tortakool then added, “We have an excellent partner like CMKL and it is a good place to start, our very own partner. One of the reasons we partner with the university here is to drive innovation in Thailand.”
Anthony added, “Businesses really need to focus on the end result, what are they planning to achieve from using AI, and then make sure your organization is ready for this change.”
Mr. Michael concluded that while AI is good or making decisions faster in the financial sector, it is also very useful for promoting fairness and accessibilities of technologies to bridge disparities and achieving positive impact for the society as a whole.