The playbook for becoming AI native

All right, well, first of all, thank you for joining and welcome to another series that we're hosting for @microONE showcasing practices, ideas, frameworks with some of the people who are really shaping the industry. Of the several sessions that I have moderated, this one feels special for a number of reasons. One is on February 5th, which is the official date of the launch of our joint book together with Terrence, Danny, and Paul. Our book, Becoming AI Native, is becoming finally orderable today. And we appreciate that even prior to the official launch date, people were pre-ordering. And a couple of days ago, I was sharing that we were number 1 on Amazon as new releases. And that always feels very good. It's not the first time that I think we're kind of flavoring the benefit of getting a ranking on Amazon. And it always feels special. The second reason is like we are literally connecting across time zones. So Sydney set us up from California. Danny and Terence are in the UK. Paul is in, is in Hong Kong and I have just landed in Dubai. And so we literally are covering several regions at the same time. As always, I think this is what we feel when we are writing our group chat, right, that we have to chase where we are. That said, maybe we want to spend this time to, to get an understanding of becoming AI native. And now I'm going to start from you, Danny, because the very first time that we kind of had this conversation on AI native that then led to the AI Native Foundation and all of that was with conversation with you. And you've been kind of a serial entrepreneur in the field for many, many years. And so you probably have seen becoming AI native in the making from the ground up. So imagine people that are wondering what is AI native, how would you define that? And that has inspired a lot of the work we've done on the book. Then I cannot hear you.

Yes, yes.

Now, yes, I mean, you heard the question, right?

Yes, I heard the question. Thank you, Mark. To make the questions to answers easier, AI Native is basically how we built the applications from ground up from different, totally different from traditional software perspective. The way how traditional software works is that you have to have the variable input in order to generate a definable output. That is basically how—

I think, Danny, again, we're not hearing you for some reason.

Can you hear me? Yes.

Yeah.

Can you hear me? Yeah. Sorry.

Yes.

Being AI native is that is that the way how AI works is that it's able to recognize whether it's recognizing your face, a picture, a video, a voice, a text, and hence you're able to actually complete a lot of the tasks in real time in a much faster way at scale. The way how we build the applications, the way how we design the business model will have to be totally different. This is exactly the same as, you know, if we refer back to 10 years ago, 15 years ago, you know, when we move from the web application to a mobile native application, the way how you design a website and the way how you design a mobile application has to be different because the user experience is totally different. The way how you design a website is from the point of how you sit in front of a laptop. The way how we design a mobile application has to be mobile-friendly where you are able to actually use your hand.

And then I think again, it goes on and on. Yeah, I think we heard pretty much, I think, 85% of everything you said. So it gives me enough leeway to maybe move my conversation over to Paul before going to Terence, since with Terence we have kind of a trajectory of, of works on AI that we have written together. Paul, you equally are an entrepreneur. You are the only one permanently in Asia. Um, and one of the questions that sometimes we ask is like, is this sense of AI native the same in your everyday operation as an entrepreneur? Do you find that Asia interprets this in the same way? Or do we talk about AI native, but we kind of referring to the AI labs in the US, to the AI lab scene in the UK? Do Do you find that also Asia is currently experiencing the very same sense of transformation with what we call in the book AI native?

Yeah, thank you for the question. Let me give a little bit of background. I was born in Hong Kong and I went to the States for university and then I stayed around for another decade or so in Silicon Valley. That was when big data was the new thing. The name of the game. And which of course that led to AI, because AI was part of the reason why AI became such a big thing in the past 10 years, was, you know, confluence of development, you know, the tons of data, GPU became available, the web, everything. So naturally I was kind of witnessing the taking off trajectory of AI, but Then I came back to Hong Kong in 2007 or something, and since then I've been here. Now, the AI native wave is a little— it felt a little different compared to the US and comparing US and Asia, or specifically Hong Kong, maybe not the rest of Asia. In the previous waves, like, you know, going web, going mobile, and going IoT, Hong Kong and China was very much in the midst of it. And I would say, especially when hardware was involved, like the IoT revolution, China was probably leading the way, or at least that's where the action, or large part of action was at, because you have to produce something for IoT, for hardware. AI is kind of interesting because of all the geopolitics going on, and Hong Kong for the first time is, group of China, and, and to be honest, we cannot access ChatGPT.com, uh, Claude, and all these other— Gemini, actually, it's, it's, um, we can't access them. So, but the sense of going AI native is, is very much real. Uh, I'm not sure about, uh, everyday small and medium business owners who I am right now focusing on, uh, but everyone knows AI is coming, and It has an interesting effect. I have a lot of connections back in the US, back in Silicon Valley, so I get a feel still on how US is happening. Hong Kong is quite different, but it gives me an opportunity and open view if you read the book. Basically, I talk about limitation forces creativity. And by not being able to participate in the greater arena, if you will, of AI, it actually forces me to look into, okay, so in my limited market in Hong Kong and maybe the greater area surrounding Hong Kong, what can we play as an entrepreneur and tech entrepreneur? What role can we play? So lots of introspection led me to believe, to realize, hey, actually there's a whole host of, um, of population demographics that's not being taken care of, especially in Hong Kong, but also true in the wider world. And in my case, I decided to focus on small and medium businesses, uh, where they are in Hong Kong, obviously, with— they can't access AI easily. But even in the US, back in the US, I was working at some big consulting companies, and, and the small medium companies are always— well, they don't have a collective bargaining power, they don't have a one market size, and so they're always kind of the last mile. So if Microsoft decides to do something, bring out an office suite or something, I mean, if it hits the small-medium businesses' needs, that's great. If not, you're left on your own. So I actually realized, you know, in Hong Kong and in China, there's a whole host of opportunity here, and if I focus on what's not being serviced, and help those people. They're very, very eager to go AI-native. And I think that's where I feel the difference this time between the US and Asia is.

Yeah, quite interesting, Paul. Quite interesting. It's also something that I don't think we can always distill and see from where we are. And it's also the fact that you are working with organizations on the ground that have an interpretation of AI-native that might be slightly different from what we see in London or in New York, right? Just to give a conventional places. So I'm gonna move to you, Terence. Look, this is for us X number of books together. We have, I think, a lot to celebrate. It all started with trends, then we did together with Danny the work on the AI Republic. We went back into the big thinking with the great Remo, and now we're back into AI. So how much do you think the industry has changed from when we were writing the AI Republic? And the reason why I ask you that for many years, the time, like the shelf life of the Republic was quite good. But in 2019, kind of unfolded into any one, 22, 23. So we felt that the book was kind of holding well with the passing of time, but then we felt something was changing and we had this urge to go into another writing. At first it was like a sequel and then it became a, right? What has changed, in your opinion, from when we started to work with AI to the point that we felt that we needed to really come out and voice our insights on where the industry is currently going?

I think, like, you know, at the time when we were working on AI Republic, right, with the benefit of hindsight, we have absolutely no clue where we were going, right? We just knew that, okay, there is this technology coming on stage, there is this technology becoming, gaining headlines in media. We had no clue where it is going. All we knew way back then when we started Nexus Frontier Tech, we knew that it could be a potential tool for businesses. We do need, we sort of know, guessed, what would be the first applications, what could be the first applications, but we had no idea how it would go, like how we go down the road, right? Now thankfully, you know, we have got someone called like Danny Goh who is like, you know, who was able to actually keep on pivoting and pushing. And I believe the term here is pivoting simply because especially when we had no idea where things are going, we would not be able to do as much as we are doing these days by pivoting. But it just shows that it was really, really difficult. So when, you know, when like, you know, during this book, right, you know, we talk about it when we were like, right, putting it together and like a revisit, you know, what, like what happened just a few years ago when we were looking at like, you know, like putting AI Republic together, right? There was just so much like it's just a completely different world. It's so very different. And I guess, you know, there are a couple of things that really, you know, change. One is definitely the arrival of GenAI. I would never thought that. In AI Republic, we're all only talking about predictions, and that was already miracle technologies. Now, you have got a screen chatbot that finally can answer some questions properly, or at least that can speak English properly, probably better than me, but as contentless as me. So that is a big change. And I think another change that we have definitely seen is The fact that there are just a lot more people who are aware of what this technology is about. Way back then, I remember it was just almost like our job was to actually be like, what do you call it? Like, take the myth out of it, right? We have to not just demystify, we have to demythologize it, right? So it's a very different game right now these days. People in general have got a general understanding what tech, like what AI do. Again, we are still at the, you know, at the frontline, not knowing where things could actually be going. Now, this is part of exciting, like this is exciting, but at the same time, this is also frustrating. So I think the, like unpredictability, the, you know, like, you know, the uncertainty as to where things could be going, it's, you know, it's pretty much consistent. And this is what we're trying to actually do a little bit of crystallization, you know, like, you know, crystal balling with our book.

That's right. That's so interesting to hear it from you, right? The way you mentioned it is true in the average public. I think we were taking a big gamble into what we think AI could possibly become. And a question I wanted to bring over to you, Danny, because you were part of the AI Republic. If you remember, a lot of the challenges we had with Nexus Frontier Tech was some time to even onboard the clients to the idea of AI. Many times we had to basically demystify with a client what AI was, what was not. It was a big part of our conversation. And I guess today it's much easier in the sense that people do not necessarily worry too much if you're having an AI conversation. They maybe want it. And question to you, when you're dealing with clients, How different is their understanding of AI? Is it just because we have access through generative AI that we kind of distribute it? What we originally wanted to do with the Arabic public was to democratize access, right? Do you find that that has happened or do you find that it is on the surface distributed access, but in reality we still having a lot of uneven use cases, distributions, opportunity? What is your experience? When you are working directly with the clients, Danny?

Thanks, Mark.

First of all, apologies that I'm actually in a cab, but I think one of the key things when we dealt with, you know, since the early days of so-called AI from 2015 until now was that How can we actually explain what is AI? There are probably many people perceive it in a very different way. Business people perceive it from the sense of the end of the world, Terminator. The technical people perceive it from the technical point of view, which is mechanical. Before LLM was born, everything we talk about AI pretty much in mechanical, whether it's deep learning or machine learning, because you still don't feel it yourself. And when that— when—

We're hoping that it comes back, right?

The keyword of natural.

Yeah.

Can you hear me?

Yeah.

Now, yes.

Yeah. Can you hear me? Yeah.

Sorry.

Yes.

The keyword of natural language is the keyword to basically change our perception. Before that, we look at AI as a mechanical traditional programming language or traditional programming tool. Now, because we're able to access AI through, say, all these available tools and the way how it communicates with us, it's in a very natural way. Now, if you come to think about it, it still doesn't do anything magnificent. It's still, you know, whatever AI can do, whatever ChatGPT can do is still what we can do. It still hasn't produced anything that is out of the range. But what it can do is that it can basically do the same thing at scale. And that is the dynamic that prompts people to start thinking that, right, if I can look at this technology in a very natural way and I can do this at scale, that will change the whole business dynamics. Which means, you know, we are now talking about AI agents and I start receiving, you know, in the past I used to receive the RFP from clients in a very mechanical way, which is that you have to list out all the requirements, You have to prepare the proposals, which is a solution. Now, I'd rather receive requests from some of the Wall Street banks to say that, right, Danny, I need to replace, I need 3 agents to replace 3 people. And these are the job descriptions. Give me 3 agents. And we will, you know, as long as you can prove me the ROI, that's how we will take it.

Wow.

Wow.

Quite interesting because I think that what you're saying is that we have also normalized the resorting to AI to replace jobs, which I think is something we were trying to argue against in the way that is currently done as a trade-off. But I believe that's exactly what the industry is definitely sending sentiments, right? That agents are natural surrogates of labor.

And they are also, and they are also then they are also able to be just, but like this, you know, managers are now able to specify exactly what is required, right? It's, you know, in the past it says, I, you know what, guys, I've got no idea, like, you know, how the, how it could actually improve our operations, you know, what, what can we do? Now it is like, this is like, this is our operations, this is what we want to do. Can you guys do it? If not, then someone must be able to. Right. So it's so specific. It's just so specific.

That's right. Why don't I ask you, Paul, this kind of building on what Danny said, what Terence commented, right? Going back into your kind of everyday experience, right? A lot of the work you do, Paul, is also with basically hardware and software, right? It's not just software, right? And so if you kind of follow where the conversations are, where is currently taking as we talk more about physical AI. There is a significant increase of dependency on the idea of robots with AI interface. You kind of been exposed to this probably before many of us simply because we were much more dealing with services and services naturally get digitized, mainly software. And one of the things that I remember Terence and I were talking about is that in the absence of clear ROI, Many, many investors will try to go back to physical hardware, right? So what is your experience in terms of this relationship between software and hardware in your line of work in— from Hong Kong, but I guess also China, right? So it's so specific to what you see at work.

Yeah, Hong Kong, not so much. But, you know, I go to China quite a bit and they are really into robots. I mean, you probably heard about it, you probably saw NVIDIA Jensen Huang last year in 2025 in CES. He rolled out a whole host of robots, and these are really becoming commonplace in China. And I've seen, you know, viral videos that look like, you know, a female robot, female-like robot that looks almost too real, and it's almost like, is it, you know, is it a real person dressed under or something? And And the net of it is, I think, and if you listen to Jesse Huang last year, he was talking about brown spacing. He's talking about, hey, you know, NVIDIA is building this environment to train with real physics. Basically, it's a game, it's game engine, but it's real physics, it's real fires, real this and that. And in, when you were planning, or before you guys were— before any company was saying, I want to— I want a new factory that can produce 100,000 cars a year, and, uh, and, and you can design it and you can actually game out how the— you can game out how the physics works, how the layout is most efficient, and you can actually train the agents that Danny was talking about, that, hey, you know, we need 5 types of, of agents, and, and these are Someone's going to move iron, someone's going to move this, and someone's going to weld. And they can actually train it all inside an NVIDIA environment. And then they build the factory in maybe 2 months, and then it's up and running. So I don't think China is that far ahead or even close to that in terms of software. But in the hardware, I've seen their machines and robots that do talks and do flips, you know, and their drones, and of course, so they're on the individual hardware level. I think China is getting very— it's doing the Boston Robotics. Is it Boston Robotics? Yeah. It's doing those a lot better than I would expect. But if you combine with NVIDIA or the Western side of the environment, AI environment, that can actually bring out a lot of opportunities. And yeah, I guess, I guess we have to adapt to, to see how we can, can create new jobs for humans. But yeah, I see it coming. I see it happening.

Yeah, that's great.

Yeah.

So I want to kind of build on this. I know there is a question in the Q&A from Danny, and then maybe I get this question to you and then we go to Danny, right? At one point in time in the book, we kind of argue that in an AI-native era, we actually could see that the idea of disruption doesn't happen like it has happened when, for example, streaming was basically killing Blockbuster, right? But it could be the opposite. It could be that a more, I would say, obsolete model starts to basically beat a digital one, right? For some people, this is counterintuitive, right? Because they might say, come on, disruption is mostly coming from new technology that are improving the efficiency, the product, the value chain. But how do you see this eventually? Does In an AI-native society, could we see a different narrative around how we disrupt rather than the traditional way in which just better technology disrupts the more older business model? And I ask you because I think what we have in common is also the business community when we teach, right? So these are questions that sometimes people ask about the AI strategy side.

So yeah.

That was a surprise to me, I have to say, you know, when we were working on it. And, you know, the fact that One would imagine, I think we were too old, let's face it, Mark. Let's face it, everyone here, I guess, we were too old. We still remember too well what happened during the dot-com boom, right? All these digital companies came in, basically wiped out all the old ones. Who among us was never actually stepped into a Barnes Noble, like a Borders? Who among us never actually stepped into a Borders? And now, you know, it was all wiped out. And like, I have to say, it was Paul who actually like took me to the first Borders in Michigan, Ann Arbor, Michigan for the first time, right? Like in the mid-'90s. And then now it's like it's got all wiped out. So I think we are way too ingrained with our thinking that, you know, like when there is a technology, it will, you know, it will like, you know, take away, like, you know, get out of the old ones, come in the new ones. I think technologies first, I think there are two things. One is because of its adaptability, especially with algorithms that you can do so many different things, right? That allows you to implement within a company, old and new. Now the key I think is what Paul was actually saying. It's old and new, they can do it because they actually have got the financial muscles, they have got the process in place, they've got the scale to take full advantage of it. But SMEs, that's still a big question mark. And like, you know, like that's something that I think like, you know, like we do need to actually address it down the road. Second is the fact that, you know, because things are like moving so quickly, the new technologies give the incumbents, the older companies, a chance to actually turn it around. Right now, in a very, very superficially, I was actually looking at the Financial Times headlines the other day. It was saying that Walmart is now doing very, very well. Walmart, of all places. And how many times have we heard that Walmart was about to be actually killed by Amazon and all these newcomers? Right. It is still around now. Maybe it could be like it could be a non-technology related, like, you know, reason. Or reasons. But I do believe that if I remember correctly, they were very, very quick in taking on technologies. They were very, very quick in taking on online sales, chatbots now. So it actually shows that companies, older companies, all of a sudden they do actually stand a chance. I would imagine that this is going to be, in my humble opinion, this is going to be a bit of a trend in the years to come. The older ones, they will— they, they can actually stand up against the newer, like, the newer entrance.

Yeah. And maybe, maybe connects to also what Paul was sharing with us, Terence, that maybe if you were basically an asset-intensive entity that you were thinking that digital transformation would basically wipe you out, now that we are reappreciating physical, it's more about upgrading the physical with an AI interface. But I guess also different way of re— way of thinking investment that doesn't necessarily have the same appreciation of a share of a company that is entirely digital, but definitely is much more actually steadier over time. So maybe that's probably also something that, let's say, brick-and-mortar players are not doomed the same way we thought they were in the Ivory Republic, which is great. Then look, there's a question that came for you. It was from André. Who basically said, look, you mentioned that it's something about the fact AI can do pretty much a lot of things that we do with just a scale, but not necessarily things that are completely outside of our spectrum of reference. But when it's about machine learning, we have seen examples of machine learning able to recognize patterns that were not necessarily something we knew about, right? So the fact that we could get surprised by the discovery of machine learning. What's your take on that? And the reason why I ask you that, Danny, and I think you're the right person for that, is Nexus started somehow as a machine learning lab. Today we are different because the AI is different, but you've been working on machine learning for a long period of time. So what is your take on that, that we could see patterns that were being discovered by the model that we were not thinking about?

I'm going to answer these questions in two parts. Number one is that the way that differentiates human to a machine or a computer is that human— computer has a very long-term memory, which means computer can remember. You can just pull a file that is 5 years old, 10 years old within a very quick snap. You can just search for the file and then pull up the file. A computer doesn't have an active memory, which means computer cannot innovate, including the LLM. Whereas human has a very short-term memory. I can't remember what I ate a week ago. I can't remember, you know, when I woke up yesterday, when I woke up this morning, which hand, which feet did I move first to get down the bed. But that, those, those those transactions still happened. Now bringing into the jobs of humans, the task that we do, that the examples that you gave, you know, to recognize patterns, those to me is a repetitive task that we can of course do if we are willing to spend the time to go into the great details to do it, to check it one by one. If you ask me to go and recognize a face of a million people, of course I can, but it will take me probably, you know, several months to go through the photos. It's nothing new. It's not creating new things. But if you ask a human, you know, if you ask a human to photocopy, you know, to copy a Bible, you probably have a, what, maximum 60 to 70% accuracy rate. You've asked a computer, it's definitely 100%. You ask a photocopy machine, it's 100%. So what I'm trying to say is that we human beings have been building machines since the very early days of our humankind. We've been building machines to replace what we can do, but we feel it's very repetitive. We don't want to do it because it's boring, because we have short-term memory. And these machines are getting better and better, getting more natural. That's why it comes with, you know, this latest technology. It becomes a lot more natural where you can communicate with it in a normal language rather than you have to go and learn a programming language to go and send instructions and get feedback. And the trend will continue to go on and on. I'm not a biologist, I'm just a technologist. I cannot imagine a day when the machine can have itself conscious. I don't know where to begin the research. Hence, I cannot say that, you know, how can we actually get it done? Of course, I will never say never. You never know, you know, how the technology advancements is going to happen. But at this stage of the LLM, if you just play around with the LLM, nothing is new. Everything that you ask the LLM, is exactly what you would expect, that you can determine whether is it correct or wrong. You have the conscience. We human beings have the conscience to be able to say, is that right or wrong? But the machine doesn't have the conscience to say, well, it's wrong.

Yeah, Danny, I'm happy you said this. You know, since I joined MicroOne, one of the things I learned is how critical human data training is. And Terence, you did actually a period of your time in 2025 on training data as well. Right. And one of the things that we kind of started to figure out is that when the kind of economy gets saturated, the only way to create new growth is by continually training models with— The success of the industry is largely determined by humans being able to understand in which way to codify their knowledge. So it's more about that. Sometimes we don't know how to codify. Rather than thinking the knowledge is within the machines. And I think this is— it's nice that without necessarily having heard the conversation, you basically validated the saying that I've been hearing a lot. So guys, I know we have a hard stop in a few minutes. I'm going to try to go to you, Paul, and then finish with you, Taren. Interviewing together with our CEO, Ali, the CEO of Box. And Of course, we see agents becoming important. Equally true, it's always true that we see agents in need of human oversight. So when you're looking at, for example, your experience in being an entrepreneur in Hong Kong, how much of this is basically modeling of a solution like large language models? How much of this is about becoming agentic? And so you see asking same people, same question that Danny was saying, people asking you, I need an agent to do this. I'd like to know because I guess the maturity of the regions doesn't necessarily show the same needs at the same time. So what do you hear when you hear about agentic AI or agents in general? Is it becoming common as a conversation when you're working with your clients?

I can only speak for Hong Kong. I mean, China is a different beast. I mean, they're very advanced in their own way. Hong Kong, again, is a small place and it's not really known for for deep technological advancements. And particularly with some kind of embargo that, you know, we don't have access to a lot of LLMs, there's not a lot of agent development in Hong Kong. Not that there's none. I mean, we do it, we do it. But probably unlike what Danny experienced, the large companies say, hey, we want 3 agents for these 3 jobs and these are the job descriptions. I think people intuitively understand that AI can replace certain job types. They wouldn't know that we want these agents. But if people like, you know, the consultants, like people like my company, right, we would go in and we're focusing on accounting and we would basically say, hey, you know, why don't we come in and try to learn what are the bottlenecks? I think we have to— we as service providers, we have to do a little more I guess giving first, I guess. People are not really, they don't really understand, hey, what are the values? But if we go in, hey, let's do a pro bono 2 weeks, 3 weeks, a month or something. We shadow you, we train, we figure out what is most hand, manual labor. And we train our agents especially, you know, manual labor stuff. We can do small LLMs and we are not, you know, we have the capability to do that. So we have to do a few more steps to understand, design and create and develop and deploy. And when they see it, I think that's when it clicks. So we go a little further. But yeah, that's the biggest challenge.

That's so interesting.

Yeah.

So interesting, Paul. Terry, look, I know you guys have actually also a launch of the book on Insight, right? So I'm going to ask you the last question and kind of putting together Ahmed's question, I'll phrase it in a way that I think could be our final question. When we were writing the AI Republic, we really didn't know what was ahead of us. I think AI native feels different because it's also the realization that it's likely that we can't go back to where we are right now. We can only go forward, just that we don't know what the forward looks like.

Mm-hmm.

But in, in history, and think about the Concorde, just because we could, it didn't mean, it doesn't mean that we will or we should, right? In the sense, if you think about it, we were thinking about supersonic flying, but eventually we didn't, right? We, we were not able to, to probably generate enough safety about the concept. And today AI is generating a lot of concern about risks and, and data surveillance and data ownership.

Yeah.

So where do you think we're gonna go from here? Do you think inevitably we, it's just about more integration, probably more distribution, more guardrails, or do you find that at the same time as we are inevitably integrating, there's also more dissent, more, more company that maybe decide to wait before integrating AI, more companies say, well, I don't really want to invest a lot of my resources into technology yet. So do you think that the trajectory is to some degree already predetermined, it's only more AI, or do you find there are pockets of adoption that maybe will show that, look, the future is simply unevenly distributed and it will continue to be that way?

I think it will be very quite like a rather uneven, you know, I, I've worked with companies that like they, they talk about like the guardrails or the importance of putting guidelines and guardrails in place. So there are companies that are willing to ring-fence any possible damage, right? And use AI in a way that is not causing negative implications to the rest of the society. Having said that, I think in a world that is like the world where we are going now, everyone is going, not everyone, like lots of people, lots of companies, like lots of people, lots of companies companies are going a lot more gung-ho than before. They are a lot more blunt about not caring about the society, will be taking a freer approach to pushing the AI boundaries and see what kind of benefits they get. And that puts, of course, the government, like authority, in a very, very difficult position, assuming that, you know, we believe that the governments in place are there to, like, you know, with a view to protect the society and not the businesses. So even the governments are actually having different views of where things are, could be going, should be going. I think we are going to see a tension that is being pulled further and further, like, you know, in the years to come. Now, what I'm really hoping, and this is my last point, is that there will be no, like, you know, no, like, a collapse or a big crisis that eventually teaches us a lesson. But I believe that the tension, we are like, we're not like at the end of the tension yet. The rubber band is still being pulled and hopefully it won't snap.

Yeah, that's right. Gentlemen, look, I, I think it was serendipitous that we could do this launch across the world on the day that the book became available. And for everybody, we had roughly 40 people always connected. And this will also go on, on YouTube. It's mostly about the people behind the project. People see books and they kind of wonder, you know, what are the protagonists that were putting the resources around this. And it's been fascinating to have this project together. I think AI Netive was a necessary step for us to, I think, to alleviate some of our angst and anxiety about what will happen with this technology. And sometimes when we write, we can put it in perspective. It's a book and we haven't done too much promotion of the book. The idea was simply to tell you there's a project that we did and We invite you to be part of that. We do talk about frameworks. We try to provide you with useful tips on how to keep on integrating the conversation on AI at the firm level. We do talk about the unchecked issue and risks that might eventually become critical. And we also talk about— so we kind of took a comprehensive view on where we think this is. We've been thinking about AI native as part of a project. There's also a foundation called the AI Native Foundation where we've been part of, where we try to kind of, I like to use the word evangelize, right? The idea that AI is an opportunity to maximize humans more and more. It's just that we have to do it in the right way. Otherwise, the shortcoming would be just an easy discounting towards the use of technology in lieu of people. And that's not what we want to. We want to have this kind of right symbiosis. So Guys, thank you so much for having animated this with me. I know that the night is long for you, Terence and Danny, as you guys have an event. And the night is probably short for you, Paul, since it's already quite late in Hong Kong. It's been so nice to see you. And I know we have events around the world where we keep on talking about AI Native. And I think we have a prospective time somewhere in the summer where we're going to be all together. Probably in Asia, but more to come. We'll reveal the details when we know them.

Absolutely.

So Danny, thank you for making this on your way from the airport. Really appreciate your time. Paul, thank you so much for having taken the time in pretty much the beginning of your night. And Terence, thank you for having been partner in crime with so many titles. All right.

And enjoy the rest of your— What did you have for breakfast this morning? Because I can't remember.

I didn't.

I see, I see. Okay, okay. I was on a flight, so I think the machine, like, I would not have a problem like mine. I had a little bit too much to drink last night, so that's the reason.

That's right. Yeah, so I was, uh, I was empty stomach for you guys. I couldn't— that you guys are gonna have very good—

very good.

All right, very good, guys.

All the best. Nice to see you again.

Take care. Thanks very much, everyone.

Thank you.

Thank you.