Terence is a Co-Founder and Executive Director of Nexus FrontierTech, a London-based tech company specializing in the development and integration of AI solutions that help organizations save time, money and resources by tacking process inefficiencies and data waste. He is also a Professor at the London campus of ESCP Europe Business School.
He is a co-author of The AI Republic: Building the Nexus Between Humans and Intelligent Automation, an international bestseller on Amazon. He also co-wrote the best seller Understanding How the Future Unfolds: Using DRIVE to Harness the Power of Today’s Megatrends. The framework contained therein was nominated for the CK Prahalad Breakthrough Idea Award by Thinkers50, the most prestigious award in business thought leadership. The DRIVE framework has also led Chartered Management Institute’s own magazine Professional Manager in the UK to name Terence as one of the 18 new voices in 2018 that reshape management and leadership. Talent Quarterly in the US called the DRIVE framework one of the 24 trends transforming talent management in the years to come.
He also wrote the book Corporate Finance: The Basics.
#11 Terence Tse, PhD – Co-Founder and Executive Director of Nexus FrontierTech on understanding the end game of AI – Get Your AI On!
Ciprian Borodescu: I’m here with Dr. Terence Tse, Co-Founder and Executive Director of Nexus FrontierTech, professor at the London Campus of ESCP, Europe Business School, co-author of The AI Republic – an international bestseller on Amazon – and co-author of the bestseller Understanding How the Future Unfolds: Using DRIVE to Harness the Power of Today’s Megatrends. I’m super excited, and it’s an honor to have you on this podcast. Thank you so much for being here.
Terence Tse: Thank you very much for having me here, Ciprian. I’m really excited to be here. And it’s been a long time, you know, like, we haven’t got a chance to talk, so when I saw your email, I immediately jumped on to it.
Ciprian Borodescu: Thank you so much. And I have to admit, I was pretty nervous about this episode, especially since I had the privilege of being your student briefly, at the 10-day MBA program in Brasov, Romania a decade ago in 2009. To me, it feels like a different lifetime ago. Romania had just entered the European Union and there was a lot of excitement, as I recall it. Today, fast forward 10 years, UK is Brexiting, we have the pandemic, the climate change, but we also have AI. In any case, this isn’t business as usual, as you wrote in the introduction of your book, The AI Republic. Do you remember those times? How did you experience Romania and its people back then? And how did it change for you over time?
Terence Tse: It’s a great question. I have been asked many times to look forward to what could be happening, what are the things that are happening. But I’ve never been really asked, you know, how do I look back? What are the things that came into my mind? First, who would have thought 10 years would change so much? If you were to ask me maybe between 1999 and 2009, I would already be telling you that, wow, you know, what? There is a lot of things that changed. These days, I think the changes are just a lot more dramatic. And you mentioned AI, and, gosh, you know, if we talked about AI way back then you would probably say, “This must be one of those mad professors who talk about things that could not really exist.” Now, AI is… Even my kids actually know, like, what it stands for, at the very least. They don’t necessarily know what it is, but things have changed. And I remember Romania vividly. First, because it was my first visit to Romania, and it was very, very enjoyable. And I think the European Union has changed quite a lot over the years, for better or worse. I think in general, things are actually better. But I think, one of the things that we’ve seen in the last year is that we as individuals have been empowered tremendously by technology. And so, our voices are actually heard more often. Now, the problem, of course, is the way we actually pick up our information and develop our knowledge is now no longer really from an extensive source. Sometimes we even got our information, daily diet of information intake from Twitter. So things have certainly changed in some dramatic ways, and, you know, whether it’s geographical, whether it’s sociological, whether it’s – like you were saying – climate, business, politically, it’s unbelievable. Just unbelievable.
Ciprian Borodescu: Absolutely. And even the entrepreneurial ecosystem in Romania has progressed a lot in the past years. We now have accelerators, a few VC funds that are active even during the pandemic, a few successful startups, and even a unicorn – actually, a decacorn if you can imagine – UiPath, you’ve probably heard of it. In fact, they are preparing for an IPO next year, if I’m not mistaken. And I think, like you said, it’s an ecosystem that has a lot of potential to grow not only in Romania but also in Europe, in general. But specifically in Romania, this is something that has been recognized by the US, Canadian, and Western Europe tech companies that acquired Romanian startups or opened research and development centers here. I wanted to maybe invite you to walk us through your journey from ESCP Business School, where you’ve been a professor for 14 years to founding Nexus FrontierTech in 2017. How did that come to be? And how did you make the leap? And why, out of all the things, AI?
Terence Tse: It was like a dumb stroke of luck, you know. So what happened was, you know, if we were to trace back, when I first started working in the business school, I thought it will be a traditional job where you will be doing research and teaching and everyone doing exactly the same things as they have done in the past 100 years. And so I remember one day I got an invitation from a bank and the bank was asking if I could actually come and give them a talk about what actually caused the financial crisis of 2009. That was basically like, two, three years after I’ve joined the business school. And I said to them, “Why do you need me to tell you guys these things? Because you’re a bank, you must know.”
Ciprian Borodescu: You already know this stuff, yeah.
Terence Tse: Essentially, what I wanted to say is that you guys caused this problem, so you must know this problem. But, you know, they said, “No, we don’t.” And then, at the very moment, I realized that the business school education… And if we look at your 10-day MBA program, we actually tend to divide it into silos, you know, into different disciplines. And so, one thing I realized at that point was that, when it comes to business education, a lot of the time it is still very discipline-driven. So, you know, there is no conversation between the different disciplines, like, chemistry is between different siloed domains of knowledge. So, at that very moment, I realized that the key is to be able to explain things from a different perspective, not just from textbooks or of research, but just as important at the very least, is by looking at how things actually develop, what are the causes. So, trying to focus on the more contextual stuff and try to make sense out of it. And it is exactly because I got out of this sense-making, of the context, that eventually led me and my partner in crime Mark Esposito, to start to develop courses that are actually taking an approach as such. And fundamentally, we call it Business 360 because that’s what we want to do, we want to have a 360-degree view as to what is surrounding a business.
Terence Tse: Over time, we wrote our first book on megatrends. We called it Understanding How the Future Unfolds: Using DRIVE – which is a framework – to Harness the Power of Today’s Megatrends. So, we started working on megatrends. And when I was going through these trends, like looking at the E as an enterprising dynamics – each of the letters actually stands for a trend that we looked at; when it came to the E – enterprising dynamics – I started to realize that technologies are going to play an ever more important role when it comes to changing our lives. I’m never one who is interested in technologies; you know, so forget about asking me to update my phone, forget about asking me to buy the latest gadgets; you can forget about those things because I’m never interested in those things. But it is exactly at that point that I realized that there are two technologies that are going to fundamentally change our lives. The first one is blockchain. And the second one is artificial intelligence – AI.
Terence Tse: So, at first, I picked blockchain because one, I find blockchain to be a little bit more intellectually stimulating, at least to me. And second, you know, the fact that I tried to read a book on AI and the book was totally and completely boring. It was downright silly in many ways. I said to myself, “This is definitely not something that I’m interested in.” But then I realized that there is no real application of AI out there other than a field, like domains like logistics. You know, they are putting some money into it, putting some research into it, but there’s no widespread adoption or widespread promotion with that technology. So I said, okay, let me actually switch to AI and see what are the things that can actually be looked at? And it was actually about writing… Like, I wrote an article on AI, and an old friend of mine called me up and said, “We need to talk.” I said, “Why?” He said, “Because I’ve got an AI company, but it is based in Asia and we want to actually roll out our services and businesses in the West. Would you come and join us?” So to cut the story short, this is how I actually made the transition to AI, and this is how I actually started. You know, like, I co-founded the Nexus FrontierTech, which is the larger entity of a previously Asia-focused AI outfit. So this is my journey.
Ciprian Borodescu: This is interesting. So the company now has headquarters in London?
Terence Tse: Yeah.
Ciprian Borodescu: Yeah. And Hanoi as well?
Terence Tse: Now we have an office in Singapore as well.
Ciprian Borodescu: Oh, excellent.
Terence Tse: What we have found is, sadly to say, Asia is a lot more advanced when it comes to using AI. A lot.
Ciprian Borodescu: Yeah. And I think we’re gonna talk about that in just a minute. But before that, in your book, The AI Republic, you talk about the acronyms FAANG – which stands for Facebook, Apple, Amazon, Netflix, and Google – on one hand, and on the other, it’s BAT from Baidu, Alibaba, and Tencent. And are these still valid today? Or do they need some sort of adjustments? For one, I think we’re missing Microsoft somewhere in there. But, do you think that the gap between AI tech giants and AI challengers has widened dramatically in the past years? Or do you feel more optimistic about the danger of AI monopolies and trusts? For sure, this is another topic that is popular nowadays.
Terence Tse: I agree with you. Yeah, it needs to be updated. We need to actually throw Microsoft back in. To start with, this term FAANG was not my invention. It was a term that was used generally by many people way back – when I say way back, it sounds like historical, but we’re talking about three, four years ago. And way back then, Microsoft was still trying to figure out what its footing is in the tech space. It’s been big, but it was always occupied doing exactly the same thing, and then the Window phones were not going anywhere. It’s only now they started to actually really pick up now. Today, they’re even talking about buying Tik Tok, right? Whether it’s right or wrong. But the idea is that we need to add Microsoft in.
Terence Tse: Other than that, I think the gap between AI tech giants and AI challengers probably has widened dramatically and will continue to widen. If you look at the stock market today, what would you find is that because of the pandemic, a lot of companies their shares basically plummeted. You know, they all basically fell into more negative regions. The only exceptions are basically the tech giants, and they’re all 20-25% higher than before the pandemic. At the same time, let’s not forget, since I’m doing finance, so I need to talk more about money, the fact is that they have got a lot of capital that they can use to invest. They can keep on buying and what we have seen in the tech space is that tech giants have been buying a lot of other smaller companies. If you look at companies like Facebook, Apple, Google, at the very least they have actually been buying. And Netflix, Amazon, and Apple, they have been investing because money is cheap. And if you think about it, when these companies actually invest, they have a slightly higher chance of success. So I think in the coming years, what we’ll see is that they will continue to dominate the space and they will continue to be able to accelerate their development, accelerate their ability to go forward. So, now, whether they are going to have monopolies, you know, I think they have already got a lot of monopolies. You know, if you just look at the pandemic, basically, how many billions did Amazon do, right? And Amazon actually has got a lot of options in expanding into different spaces, like different market spaces as well. So I think the monopoly is already taking place. The question here is, you know, should we be trusting the tech giants? The short answer is, I think many of us, we’ve got no choice. You know, like, can we actually stay away from our phones?
Ciprian Borodescu: That’s a good one.
Terence Tse: Can we not buy things online? It’s difficult. And, as a matter of fact, whether you buy things online, whether you use your phone, wherever you go, data is collected all the time. So, you know, how companies, in general, are going to use data is going to have an enormous impact on the level of trust that people would give to them. And there is a recent book called Surveillance Capitalism. It’s a very, very thick book. But in that, it talks about, in the author’s view, how Google actually were able to make money out of data. And what she was trying to say was, if you were to use Google, it is not just that the Google search engine is now 80-90% of the market share. That’s not the ultimate goal. What they want to do is to actually collect as much data as possible, so that they can actually come out with very accurate predictions, which they sell to marketers. And if you are any company that is trying to do marketing and someone can sell you fairly accurate predictions, it is almost like having a money printing machine sitting right in front of you. So, it’s like, are companies going to be able to hold back with the use of data, with the use of private data? I don’t know. So, it’s a battle that is still being fought in many ways.
Ciprian Borodescu: And I guess that’s where governments could – and maybe should – intervene?
Terence Tse: Yeah, yeah. I mean, if you look at the very, very painful thing called GDPR – like, in Europe, General Data Protection Regulation, right? As painful as it is for many companies, it is actually a good thing because this is the first attempt – basically in this case set up by the EU government – to actually restrict the private data to be collected and to be misused. So that’s a very, very big step. That’s a very, very important step. The question here is, with the US or other parts of the world, would they be happy, willing, or able to come up with something very, very similar? Because if you think about it, if EU is the only place that is doing that, it is going to place EU at a disadvantageous position when it comes to already behind catching with AI.
Ciprian Borodescu: You’re painting a gloomy AI present or future and, to be honest, I’m also helping you a little bit here, but I’m just wondering, at the same time, how can AI startups and companies, like Morphl or Nexus succeed with limited access to funding and talent? I mean, what are the levers that we can pull to push through and reach a point where AI is no longer considered elitist? Because this is basically what we’re dealing with today, right?
Ciprian Borodescu: Good question. Yeah, that’s a great question. So, one reason why we are operating in Vietnam is because labor is cheaper – the staff is cheaper – so we can afford to hire more staff. But even in a place like Vietnam, you will be surprised how high the salary level is already. It’s still not quite at the US level, but the rate it is going up is absolutely crazy. If I’m not mistaken, they are looking at salary increases twice in a single year. Twice! The gap is actually closer. But I think there is a general misunderstanding when it comes to AI talent, and this, in many ways, deters many companies from taking on AI – this is an observation. And the fact is this: when it comes to AI talent, people, when they talk about AI talent, they tend to think about the AI scientists, the people who actually build the core AI model, which is fine, which is quite important.
Ciprian Borodescu: But that’s one component, right?
Terence Tse: That’s only one component. Exactly. That’s only one component. There are so many things like, you know, how do you actually make sure that your AI model, for instance, would work in the existing IT infrastructure, right? In that case, in order to turn whatever the model can generate, you need AI engineers, developers. Now, these are the people who we found to be a lot more important. So the first thing is not to focus on AI scientists. Now you may be asking, without AI scientists, how can we actually build a model to start with? The answer here – I believe the biggest obstacle when it comes to businesses is putting AI into their businesses – is that they always tend to think that it is something that is rather complicated and sexy. Ultimately, for all businesses, when it comes to using sophisticated technologies like AI, is asking the question, What exactly do you want to actually achieve?
Ciprian Borodescu: Yeah.
Terence Tse: If what you’re trying to achieve is something as simple as reading documents, you don’t really, really need to have an AI scientist. You probably need to actually buy an AI model or get a vendor who can come up with a piece, like an algorithm that will allow you to achieve that purpose. But you don’t really need to have your own AI scientist. What you do need to focus on, again, is the engineers, as well as the business people who can run these engineers or at least deal with the vendors. So, I think there is a lot of misunderstanding when it comes to AI as a technology, when it comes to AI as a business proposition, and when it comes to AI as human resources.
Ciprian Borodescu: Yeah, I think that’s a fair point. And from a digitalization point of view, I think it’s almost as if we’ve jumped from 2020 to 2030, in a matter of months if we take into consideration this pandemic, and I’m still trying to figure out the implications on that. On one hand, if we take the retail industry, for example, clearly more and more businesses have today an online strategy versus a year ago. And, at the same time, more and more consumers are buying online. And long term, these trends will hold. But on the other hand, I can’t help but wonder who are the real winners here? And what can we learn from all of this? From this forced pivot that took a lot of businesses by surprise. And it kind of pisses me off that it took a pandemic of this magnitude to force companies to prioritize digital transformation. How crazy is that? I mean, do we need to go through a similar painful experience for companies and people to seriously adopt AI?
Terence Tse: As much as media talks about companies using AI to do this and that, this and that, what you will find a lot of time is they will actually do the AI this and that, this and that, but only up to the proof of concept level. And proof of concept basically means pilot, but a lot of them stop the pilots for whatever reasons, you know, like, they can’t pull it in the assistance system, or no one has got the interest to carry over. So, unfortunately, there has been a lot of inertia. There’s a lot of complacencies, if you like, when it comes to using AI, so when it comes to not doing anything technology. Besides, it all depends on what exactly you’re trying to achieve, again, because if you have got a choice between using digital technologies forward, but if you’re hiring people to do it is still cheaper, why would you bother? You know, the return on investment is not there. The proposition is not there. So, now all of a sudden, things are very different, because, for instance, in the past, you may be having a room full of people who do nothing but document processing. Now, because of social distancing, you can’t have that many people in the same confined space, all of a sudden, essentially you see your document processing capabilities cut by half, right? What are you gonna do? So, all of a sudden, you have got no choice but to actually use technologies of any sort, including AI to do that. But we’re human beings, right? Until something that is big and bad happens, we like to stick to a stable life and stability.
Ciprian Borodescu: Yeah, it has to be painful to change, right?
Terence Tse: Yeah, exactly. And the change is painful.
Ciprian Borodescu: Can you give us some examples of projects and products that you’ve built at Nexus FrontierTech? And what were the biggest challenges? Both from a technical point of view, and also business. And how did you overcome them?
Terence Tse: Sure, sure. So, we are a small company, but we were lucky enough to actually have a global bank as our major client. And what we do is something very simple, in this case, anyway. We literally just read their documents. That’s it. So you may be asking, What’s the big deal about reading documents? At first, I thought, like, gosh, you know, document reading is so terribly unsexy, so what’s the point? But the point here is this: two points for our clients, you know, when it comes to helping them to read documents. And by the way, we’re not just talking about reading the documents on PDF, we’re also talking about, when you go and apply for a mortgage, you need to submit different types of documents – from driving license, ID, to your bank statements, to your payslips, that sort of things – and, what we do is literally, we have a technology that can help our clients to read all sorts of documents, regardless of what format they’re in, and pick this data from these documents, which we call unstructured data because they are not on the database. Even though you have got the documents yourself that are sitting on your hard drive, the fact is that the data itself is still unstructured because it’s not prepared for use. So what we do is we extract and then we prepare them and we put them on to the database, right? That’s basically all we’re doing, using algorithms and natural language processing. But the key here is that all of a sudden, it helps the bank – not cut the cost; in this case, we’re not really helping them to cut costs – more in the line of preventing future addition of cost. Because, instead of having five people doing the work, we’re not saying that, oh, you can now only have three people doing exactly the work of five people – what we’re doing with them is, you have five people right now, but what happens later on when you increase the volume of your business? You can now have five people doing the work of eight people that use the tech.
Ciprian Borodescu: Basically efficiency, right?
Terence Tse: Yeah, exactly. The efficiency is going to the future. We’re not trying to extract it from what we have got right now. That’s one. And two are the benefits. We help the bank to be more able to comply with regulatory requirements. Because when it comes to human checking it, is very, very slow and expensive, and they make a lot of mistakes. And over time, what we’ll see is that the governments will ask all the banks to be more and more compliant. So, it is also about risk management for our clients. When it comes to challenges, it’s always the same thing. People. You can’t find people. You know, we always say this: there are basically two types of people we need to actually make it happen. One is tech people and the other one is business people. Because after all, if you want to put technology into a business, you need both types of people. Yet, you know, tech people have got very little understanding or have got no interest in the business side of things. Business people don’t know much about technologies, and therefore, they try to stay away from them. And you almost always have a gap between the two.
Ciprian Borodescu: Like you said in the beginning, these two kinds of people don’t need to be in silos, because they have to communicate between them, right? The technical and the business people.
Terence Tse: Correct. Correct. Correct. And you do need people like this in order to facilitate the implementation of the digital transformation. Let’s face it, this is the only way you can get it done. So, that’s why I always say to anyone, any young people who are willing to lend me their ears, I will say, you know, go and study business, that’s fine. But make sure that you know something about technologies because, in order for you to make the most use of your business skills, you do need to know how technologies work because this is the area where you will be able to excel.
Ciprian Borodescu: Excellent. I want to pick your brain on the diversity topic – where do you see these things heading? Personally, I feel like there are more and more women involved in STEM, which is amazing. However, even when trying to hire people at Morphl, it still feels that we need to be intentional about it and be on the lookout for women that are interested in a technical career. And I’m just wondering how things are based on your experience? What’s the ecosystem in UK or Asia versus Eastern Europe?
Terence Tse: It’s an extremely important question. So I read quite a few cases where women are good at doing the jobs in tech stuff, working at tech companies. And you know what happened? The type of discrimination they receive and face day in day out is enormous. And there is harassment coming from men, you know, uninvited comments, like deliberate actions. In many ways, right? At the same time, when the government is trying to promote them, but they’re not doing anything to get companies to be more accommodating to women’s needs, you know, sometimes they’re saying, “You know, you’re going to have your baby and therefore, you know, we can’t do it.” So, it’s almost like, what’s the point of governments pushing for women to be more involved in STEM? Don’t get me wrong, I think it’s a wonderfully important thing, but if we’re not changing the employment side, the demand side of things, just by changing the supply side, it won’t work.
Ciprian Borodescu: That’s a good point.
Terence Tse: So, at Nexus, this is one of the reasons why we’re very, very proud of ourselves. Because at Nexus, we have got 50-50%, when it comes to male, female. We’ve got a lot of females working in our company. As a matter of fact, all the people who are calling the shots at the operational level, they are all women, and we’ve got a lot of talented female engineers, too. So, it makes a huge difference when it comes to not just diversity – diversity, of course, if you have 50-50 you have good diversity – but it also makes the culture a little bit more different, it runs a little bit smoother too. It’s a lot more dynamic in many ways.
Ciprian Borodescu: Yeah. And it provides different perspectives, basically, in the business stuff, right?
Terence Tse: Yeah. And I would recommend any companies to push for that, not to just treat it as lip service. You know, it’s about time to stop the lip service in anything.
Ciprian Borodescu: Yeah, yeah. And talking about PR, another point is that, you know, clearly AI is hyped these days. And this is something that you recognized also in your book, and I would like to invite you to give our listeners a framework or a set of questions that can help them identify those companies or startups that are indeed doing AI versus those that are just playing around with some keywords on their landing page.
Terence Tse: Yeah. Okay. So yeah, we all think AI is the solution to many problems. No, it’s not true at all. What AI is, these days, is nothing more than an algorithm. So, if you’re gonna think an algorithm is x plus y equals z, that’s basically what it is. So if that is what all AI can do, what you need to do is to actually think in very, very, very narrow terms what exactly is the task – not activity, the task, the business task – you want to accomplish? You want a solution. So if the task is all about reducing the number of people who need to go through documents, that is a task, right? If you’re trying to actually reduce the number of people, like labor hours, for a business unit, that’s a different story. Because ultimately, if you do not know the ultimate goal that you want to deliver, no AI will be able to help you out. So think very, very, very narrowly. The second thing I will always ask everyone to think about is, you know, don’t even think for a moment that AI is going to do all the work for you. No, it’s not. You always, always, always need humans to do some work – those things that machines cannot do. And this was one of the reasons why the five people we have got for the bank to do document checking, they’re gonna stay there. Because what happened was that we built a model that can check UK documents. But what happens if there is someone who is Japanese, submitting a Japanese passport or documents from Japan, right? So we are not going to build a model just to cater to the Japanese clients, because there are not a lot of them. So, these exceptions will all be left for humans to do. In other words, what we’re doing is we use machines to do most of the work, but we use humans to deal with the exceptional cases or things that machines cannot do.
Ciprian Borodescu: And for those that want to understand the end game for AI technology, I strongly encourage them to read The AI Republic – it’s on Amazon. And honestly, there are many things that resonate with me from the AI versus IA analogy – big data versus right data – or the MAP – minimum algorithmic performance. I think these are important notions and concepts to be understood by any leader out there that wants to incorporate AI into their organization. What are three actionable takeaways you’d want entrepreneurs or executives or managers to remember after reading your book and immediately be able to apply to their startups or organizations?
Terence Tse: Okay, okay. I think the first thing that one needs to understand is this: the most important part of AI is not the model itself. In a couple of articles that we wrote, we always use… If you think about AI, essentially an AI model, a good AI model is like a performance car engine. Like, you can have a Lamborghini Ferrari engine, but you need the rest of the Ferrari, you need the rest of the Lamborghini in order to go from point A to point B. So, the key is actually not to focus on the AI model itself. As a matter of fact, lots of companies actually buy AI models from the other vendors, like are now because there’s so many AI developers. Again, you know, it all depends on what you’re trying to achieve. If you’re trying to actually go and try to beat a deep mind when it comes to goals, that’s a completely different story than, you know, you’re trying to use AI to deal with customers’ questions online, right? So, it’s a very, very different thing. So, think not about the AI model itself, but the rest. The key to success is to get the rest correct. You can do it yourself or you can actually get vendors, to work closely with vendors, to make sure that the so-called production environment is actually up to scratch. So, think about that, because no matter how many PLCs and pilots you do, you know, if you have not got anyone to help you to implement, you won’t be able to handle it. The second actionable takeaway I believe is, you know, do not even think that you need to collect enough data in order to actually run AI projects. Because, again, what really matters is what exactly you are trying to achieve.
Ciprian Borodescu: The use case, yeah.
Terence Tse: So, let’s say, if all you’re doing is actually getting a model to make sense and read and understand and extract the data from a driving license, that’s easy. Okay? Because on a driving license, the data is always in exactly the same place. Whereas, you know, if you’re asking a machine to actually make sense out of bank statements, or statements coming from different banks, they all have different formats, that’s a slightly different story. Now, what we’re doing these days is, when you do not have enough data, you can already start to synthesize it, you know, we create for you, right? Now, it is not terribly difficult, because all you need to do is to have a certain set of data to start with, and then what we’ll do is we’ll make some changes to the data that you have got – the current data you have got – and then we’ll push it to actually train the model to be more accurate or to be more able to read a document.
Ciprian Borodescu: And, of course, it’s not perfect, but at least it’s a starting point, right?
Terence Tse: Oh, it’s not perfect. You came up with a very, very important point. A very, very important one. Do not even think for a moment that it will be perfect. You know, like, if someone is telling you that you can get 100% out of it, the person is lying to you because it’s not gonna be 100%. And you have to actually understand the fact that you will never be able to get to 100%. So, my third takeaway is this, you know, think about what percentage is actually acceptable? For a lot of things you don’t need 100%, right? If you are just checking documents, you don’t need 100%. If you are using AI to diagnose cancer, probably, you will want to look at 100%; or you’re using AI to decide whether someone should go to jail or not, probably, you’re looking at the 100% accuracy. But in many cases, you don’t really need 100% accuracy.
Ciprian Borodescu: Absolutely. For example, in marketing or in sales, you can go with 70 or 80%, something like that.
Terence Tse: Exactly. So if we were to run with what you were saying, like, let’s say even 70% is acceptable, right? Now, the question here is this: if I come in and say, “Hey, Ciprian, I will be able to get you to get to 70%.” I may be lying to you. Do you know why? Because, without ever seeing the data you have, how can I actually tell you that I’ll be able to bring you to 70%? So what it means is that if you are getting someone to use AI for your business, you need to be able to keep on adjusting what is the acceptable level of accuracy over time. Because without the first trial, you won’t be able to get to the end game. And so, you know, it’s a bit of a chicken and egg issue, but it solves a lot of problems if you are happy to live with a circular reference situation like this.
Ciprian Borodescu: And if we just continue down this path, what are the main obstacles and objections when talking to the leadership of a company? From your experience, where is an AI project getting stuck? Or where things can go wrong at different levels across the organization, based on the customers that you guys have, and stuff? I’m just wondering because we also have these kinds of conversations and sometimes we hit a roadblock.
Terence Tse: I think, on many occasions, the roadblock is the process. Putting technologies into a business almost always requires the redesign of the processes, unless the process is already in place. For instance, you are trying to use AI as a chatbot. Right? So, in that case, the answering process is already there, but you still have to actually figure out what happens if the chatbot cannot answer the questions. You know, what is the redesign of the process? And I couldn’t believe it myself, but what I’ve seen is that many companies do not actually know the processes. The companies will have to overcome themselves before they can start talking about using AI.
Ciprian Borodescu: Yeah, so start with the process, understand the process, and then…
Terence Tse: Start to map out your roadmap, like, start to actually map out your workflows. That will be a very, very good start when it comes to making the first contact with any AI.
Ciprian Borodescu: This is such a good point. This is such a good point. What’s your plan, Terence, for the rest of the year? And what are the growth opportunities for your company, as we move out, hopefully, of this pandemic, in 2021?
Terence Tse: A personal goal is to survive, stay sane, not try to kill myself in the lockdown. Lockdown is driving me nuts. I’ll probably want to see more people, especially when it comes to teaching. Teaching over Zoom is no fun. But other than that, is about really staying healthy. But, as we talked about earlier on, the speed of taking on digital transformation has accelerated; as a result, more and more companies are using that, and we are now receiving more and more queries than before because everyone is doing it. So, I think the growth opportunity is actually in the digital transformation. Now, it all depends on which side of the digital transformation you’re in. If your job is being threatened, I think it’s about time to think about how to retrain yourself or do something different. But if you are school graduates, I think this is actually a good time. Even though the job market is going to be tough, I think there are tons of opportunities – like non-traditional opportunities out there – and this could be an exciting start of wonderful adventures ahead of them.
Ciprian Borodescu: Yeah, yeah. Yeah. Awesome. Terence, it was a pleasure to have you, and thank you so much for sharing your wisdom with me and with us. How can people reach out to you for ideas and comments?
Terence Tse: I think the best way to do it is to actually hit me up on LinkedIn. I tend to actually use LinkedIn more often now than anything else when it comes to getting in contact. So yeah please do that. Please use my LinkedIn – it’s Terence Tse, PhD. Ciprian, I want to say thank you very, very much for your invitation. It’s great to hear from you again. The pleasure I’m sure it’s all mine. And thank you so much to all of you for spending almost 45 minutes listening to me. I hope you got some benefits, some value.
Ciprian Borodescu: This is awesome. Thank you so much for your time, Terence.
Terence Tse: Thank you very, very much, Ciprian.