A conversation with Manus AI's cofounder and CPO Tao Zhang

Transcript

AI isn't just moving fast. It's moving at a pace we've never seen before. At Stripe, we're seeing the data firsthand. AI startups are rewriting the playbook and scaling to revenue milestones years faster than the giants that came before them. But how are they doing it? And how do you sustain that kind of momentum? In this series, we sit down with the founders behind this historic rise to find out. This is the AI boom on Stripe. Tao, thank you so much for being here with us today. Yeah, me too. So I'd love to take us back to the first time that I learned about Manus and I think probably the whole world. So it's March 2025. We've had AI chatbots for two, three years. Everyone's been interacting with chat, but that's really it. And then a video starts circulating the internet. No flashy stage, no Silicon Valley keynote, just a raw screen recording of your product. Yes. Tell us what this video shows the world. Yeah. I think it's just seven days before the launch date. And when I'm looking at our website, I think it's too simple. So I was thinking about maybe we should make a video for our launch. So I call some video agencies and they all told me that it took them maybe two weeks or three weeks to build a launch video. And I told them that we only have six days. And they said, oh, this is impossible. So we decided to shoot the video by ourselves. Yeah. So I'm using my own camera. I shot the video and Pete, our chief scientist, is the actor there talking about the product. So I shot it and I edited it. Yeah. And because we're not a very professional launch studio team, we only put Pete in the camera for maybe one minute talking about our product. The rest of the time is Pete demoing our product for three use cases. I think that can be the main reason why we went viral in the first place. Because we just show the world what an AI product could be, other than a chatbot. We actually have that video right here. So let's take a look. Hi, I'm Pete from Manus AI. For the past year, we've been quietly building what we believe is the next evolution in AI. And today we're launching an early preview of Manus, the first general AI agent. This isn't just another chatbot or workflow. It's a truly autonomous agent that bridges the gap between conception and execution. While other AI stops at generating ideas, Manus delivers results. We see it as the next paradigm of human machine collaboration and potentially a glimpse into AGI. Now, let me show you Manus in action across three completely different tasks. Let's start with an easy one. In this example, we'll ask Manus to help screen resumes. I've just sent Manus a zip file containing 10 resume documents. Since each Manus session has its own computer, it can work like a human. First unzipping the file, then browsing through each resume page by page, and recording important information to documents. Manus works asynchronously in the cloud, which means you can close your laptop anytime, and Manus will notify you when everything is complete. So the video shows basically that with Manus you can't just chat, you can actually execute on tasks. [SPEAKER_01] And it writes Python code and it goes in Opus browser. [SPEAKER_03] Yeah, does everything. [SPEAKER_01] And it goes entirely viral. I think you had 2 million signups for the private preview after that. [SPEAKER_03] Oh, yeah, yeah. We have 2 million in the waitlist in the first week, and we got 3.5 million in the first month. A lot of people on the waitlist. [SPEAKER_01] And I think I read somewhere that people were reselling the invite codes for up to $14,000 online. [SPEAKER_03] Yeah, at that time, if an influencer got an invitation code to Manus, they would attract much more traffic compared to the price they paid for the invitation code. But we didn't get any money for that. [SPEAKER_01] Yeah, no, that's a whole new topic. We could do a whole different podcast on the internet now with AI, but we'll leave that for a different time. So what do you think was it in the video about the product that struck such a nerve with people? What was the aha moment for people at that time? [SPEAKER_03] Yeah, because I think before Manus, people treat AI just as an answer machine. I give you a question and you give me an answer. It's telling you how to do things, but in the end, you have to do things by yourself. But in our launch video, we always focus on the outcome, on the deliveries. Maybe it delivers you sites, it delivers you a website, or it delivers the actual code that works. So I think people just get amazed. And another thing is that in the whole process, when the AI is doing things, it's totally without human intervention. You just tell Manus what you want, and Manus will figure out what's the plan and how to execute the plan, step-by-step all by itself. It's all automatic. So I think that's really something new and amazing for the whole world to see. [SPEAKER_01] Yeah. I remember thinking, oh, this is the first time where I felt, oh, this can actually save me time and help me do my job. Yeah. Versus just a little bit of a thought partner. [SPEAKER_01] Yeah. It was really magical. Did you know it was going to be so successful? Like the night before launch, where you're thinking, yes, we've hit on something really special? [SPEAKER_03] Yeah. [SPEAKER_03] You know, we are very small staff at that time. Yeah. We only have 40 people there. And we thought it's a very good product because when we developed the product, we will get amazed by our product just every day. So every day in our office, it's like a feeling of, oh, it's awesome. Oh, how can I solve this? Every day is somehow like this. So we think it's going to be successful. [SPEAKER_01] The night before launch, where you're like, yes, we've hit on something really special? Yeah. We are very small staff at that time. [SPEAKER_03] Yeah. [SPEAKER_03] We only have 40 people there. [SPEAKER_03] And we thought it's a very good product, because when we developed the product, we will get amazed by our product every day. [SPEAKER_03] So every day in our office, it's a feel of, oh, it's awesome. Oh, damn, how can I solve this? [SPEAKER_03] Every day is somehow like this. [SPEAKER_03] So we think it's going to be successful. [SPEAKER_03] But I think nobody will expect that we can be at this level of successful, yeah. Yeah, that's a high bar to clear. [SPEAKER_01] Yeah. [SPEAKER_01] So this is a good time to talk about the product, I think. [SPEAKER_01] I think from the very beginning of Menace, I feel like you really reinvented what agentic AI products should look like. [SPEAKER_01] Yeah. [SPEAKER_01] So I wanted to go over some of these product areas with you and talk through some of the thinking that you had behind them and why you developed them. [SPEAKER_01] Yeah. [SPEAKER_01] Virtual sandbox, tell us what is that? The virtual sandbox, but internally we call it a virtual machine, yeah. It's super important as it's going to be the fundamental part of the whole Menace system. Because for us, we think in the past why most AI wouldn't work is because they only use the brain. Yeah. So maybe here is a time I can share why we choose the name Menace for this product. Yeah. We take this word Menace from MIT's motto, which is Menace et Menace. It's an old Latin word, which means man and hand. And man, we think they are LLMs, they are the models, they are super smart brains. But only with brain, us humans, we can make real impact to the physical world. That's why we think we have to be the one who build hands for this very smart brain. Then they can operate the tools and they can make impact into the physical world, right? So I think what wouldn't work for other AI tools is because they only have the brain. So what they do is they keep thinking, keep thinking inside the head. [SPEAKER_03] But what we do is that we provide a computer to the AI. [SPEAKER_03] So anything the AI thinks, maybe I should do something, or maybe I should run some code to test my thesis. [SPEAKER_03] It can use its own computer to browse the internet, to run the code, to store a file on the file system. [SPEAKER_03] Or maybe 30 minutes later to retrieve the document again from the file system. [SPEAKER_03] So the virtual machine really changes everything because it provides the most powerful tool, the computer, to the AI. [SPEAKER_03] Yeah. [SPEAKER_01] Now you're building tools for the AI to do its job. [SPEAKER_01] And that's really, really incredible. [SPEAKER_01] One of the interesting product decisions that I thought you guys made was for the user to be able to see what was happening. [SPEAKER_01] Like opening the browser, writing the code, etc. [SPEAKER_01] What was behind that decision? Oh, yeah. Actually, that's the first important decision we made after we kick off the project. Because we saw that agents are so new. And people are not very familiar with this kind of product. And so we think it's very important at this stage, maybe it will last for one, two, or three years. [SPEAKER_03] And so the first thing that we did is to show every step, the whole process to the users, to make users understand what is happening in the background. [SPEAKER_03] And then they can trust the product. [SPEAKER_03] Yes. [SPEAKER_03] This is very important. [SPEAKER_03] So we build transparency as a feature in our product. Yeah. [SPEAKER_01] It's an incredible feature. Tell me a bit about the async workflow. [SPEAKER_01] Yeah. Because you create that trust. [SPEAKER_01] And now the user can step away and not really worry. [SPEAKER_01] Where did that come from? In our founding team, we're all engineers. Yeah. Even I run the product team, but I write code for more than 30 years. Yeah. So we all know coding. So we're all Cursor fans. We use Cursor a lot. But there's a problem, yeah, that when you are using Cursor, you have to keep your laptop open. Because it runs locally on your computer, right? So you have to keep it open and watch it. Yeah. Because in the first maybe four or five months, it needs us humans to click the accept button. Yeah. Sometimes it will ask you, accept, accept, accept. Yeah. So when we use the product, we say, oh, why does Cursor need me to click the accept button? It's because it runs locally on my computer. So it needs my permission to run these commands. Or it has a chance to break the whole computer, right? Yes. But in Menace, the computer is in the cloud. It's not your computer. So it's super safe to perform each step just by the agent itself. So we don't ask the users to keep pressing the accept button. Instead, we just let the agent run everything in the cloud. Then it gives us another advantage, which is the async advantage. Because everything is running in the cloud. Yes. Which means after you assign your task to Menace, you can just leave it. You can close your laptop. You can put your phone back in your pocket and do your own things. After maybe 10 or 20 minutes, when the work is done, we will give you a notification. And you will see the outcome. What do you use Menace for? Oh, a lot. Nearly everything. For me, I'm doing prototyping every day with Menace. So whenever I have a new idea, I will not just ask my designers to put the ideas alive to me. I will use Menace. Instead, I will just tell Menace everything that's in my head. And Menace will build the first prototype for me. And this is not just something we prototype in Figma. And you will see the outcome. What do you use Manus for? Oh, a lot. [SPEAKER_03] Nearly everything. [SPEAKER_03] For me, I'm doing prototyping every day with Manus. [SPEAKER_03] So whenever I have a new idea, I will not just ask my designers to put the ideas alive to me. I will use Manus. Instead of that, I will just tell Manus everything that's in my head. And Manus will build the first prototype for me. And this is not just something we're prototyping in Figma. This is a fully functional prototype, which means it has the front-end, back-end, and it has the AI capability inside that. So in this prototype, you can just try and play with it and tweak it. After the prototype is done, I will give the prototype to my designer and engineers, and they will instantly understand what I want. Because this is the real thing. Yeah. [SPEAKER_01] That's putting the fun back in software development, I think. [SPEAKER_01] Yeah. [SPEAKER_01] Who is your ideal Manus user? [SPEAKER_03] Right now, I think we describe them as non-tactical users. [SPEAKER_03] Yeah. [SPEAKER_03] Because in the first months when we kick off the project, at that time, we already have Cursor, Windsurf, Devin, all these fancy tools for engineers. Yeah. So we decided that we will not build another fancy tool for engineers because they have enough. But for normal people, for average users, they are underserved. Yeah. They don't have these fancy tools. [SPEAKER_03] They don't have cloud code. [SPEAKER_03] They don't have Codex, right? [SPEAKER_03] So the way we build Manus is to build Manus for everyone. So the ideal user is someone who wants to achieve much more with AI, but with less technical background. [SPEAKER_01] Yeah. [SPEAKER_01] I love that. [SPEAKER_01] And there's so many more of them. Yeah. So many more of them. That's it. A bigger market. Okay. [SPEAKER_01] You've been coding for 30 years. [SPEAKER_01] Yeah. [SPEAKER_01] So what do you think, is it different building for AI products or do you think really fundamentally software is software? Oh, I think building AI software is totally different. For me, I'm already in this industry for more than 15 years. Yeah. And before, the way we build product is somehow I have an idea. And I just build the prototype on Figma. And then I talk to my designer, talk to my engineer, and they will do the design, do the engineering design. And then write code. We test it. We launch it. And then we hear the feedback. And then we iterate. Yeah. That's the way we do things before. Yeah. But now I think it's going to change because previously I think product managers and designers, what their job outcome is somehow a PRD, right? Or design. It's about interface. Yeah. But in the AI world, the interface doesn't exist anymore. For most AI products, it's just about conversations, right? So I think the conversation is the new interface, which means the prompt itself is the interface. So I know there are many, many companies today whose prompts are still written by engineers or researchers. But I think a better way to do that is the product manager write the prompt themselves. Because the way you write your prompt is just the way you design your interface ten years ago. [SPEAKER_03] Yeah. [SPEAKER_03] It's the new PRD. [SPEAKER_03] Yeah. It's a new PRD. It's a new interface. So the way we do product here is that whenever we have a new idea, we will just use Manus itself to prototype with our idea. [SPEAKER_03] And after we feel the function works really well, and this is the feel we want, and then we will go to our engineer and ask them how to put this prototyping in our production environment. [SPEAKER_03] So once the engineer told us this thing can be delivered in our production environment, the last step is to go to the designer. Yeah. [SPEAKER_03] So the whole process is really in the opposite. Wow. Yeah. [SPEAKER_01] Amazing. We have the final experience first, and then we have to make it engineering ready. [SPEAKER_03] And the last step is to ask the designer, everything is ready. Just do the design. Completely change the sequence. [SPEAKER_01] Yeah. [SPEAKER_03] Completely change it. Amazing. So at Stripe, we like to paraphrase this quote from Pablo Picasso, which is when art critics come together, they talk about form and function and meaning. And when artists come together, they talk about where to buy the cheapest turpentine. [SPEAKER_03] Yeah, definitely. [SPEAKER_03] Actually, I put that sentence in my personal profile. [SPEAKER_01] Really? [SPEAKER_03] Yeah, I really love that sentence. That's amazing. We love it here too. [SPEAKER_01] Yeah. [SPEAKER_01] I think one of the reasons I love doing these types of interviews is that the builders and the co-founders of these companies, you guys are the artists. [SPEAKER_01] Yeah. [SPEAKER_01] So I want to talk real tactics now about how you've built Manus and really built the company and gone globally. [SPEAKER_01] So you're headquartered in Singapore. Yeah. [SPEAKER_01] You now have teams in Tokyo and in San Francisco. Yeah. My first question for you, when do you sleep? [SPEAKER_03] Oh, yeah. [SPEAKER_03] For me, it's a harder question because I'm the one doing all the public speeches around the world. So I really travel. So I will sleep whenever I can on the plane, in the car, whatever. But I think for most of my team, they will sleep very late in the midnight. Usually, I think our core engineering team will sleep after maybe 1 a.m., 2 a.m. That's very normal. [SPEAKER_01] Wow. [SPEAKER_01] My question was meant in jest, but I'm glad you answered that. [SPEAKER_01] You have a very hard working team. That's very normal. [SPEAKER_03] But definitely, they will wake up much later. So, I really travel. [SPEAKER_03] I will sleep whenever I can, on the plane, in the car, and whatever. [SPEAKER_03] But I think for most of my team, they will sleep very late in the midnight. [SPEAKER_03] Usually, I think our core engineering team will sleep after maybe 1 a.m., 2 a.m. [SPEAKER_03] That's very normal. Wow. [SPEAKER_01] My question was meant in jest, but I'm glad you answered that. You have a very hard working team. [SPEAKER_03] That's very normal. [SPEAKER_03] But definitely, they will wake up much later. Yeah. Yeah. Yeah. So, usually, our office starts after 11 a.m. Because most of us will come to office after 11 a.m. Because they sleep very late in the midnight. Yeah. [SPEAKER_01] And you were saying you spent 400 hours on a plane last year. Oh, yeah. [SPEAKER_01] I spent 400 hours in the air. [SPEAKER_01] Yeah. [SPEAKER_01] I'd love to see those travel rewards for you. [SPEAKER_01] Yeah. You say there are about 100 people working at Manistar. Yeah, 100 people. [SPEAKER_01] Most companies don't build global teams until they're much, much bigger. [SPEAKER_01] Yeah. [SPEAKER_01] Why did you decide to build out different offices in different locations so early on? Oh, yeah. Because after we launched Manistar, we went super viral in the whole world. [SPEAKER_03] And we have a lot of Manistar fans in different countries. And we believe it's super important to be there. [SPEAKER_03] We don't want to just be like someone in the cloud. Yeah. [SPEAKER_03] We just want to do face-to-face. So, even after maybe just 15 days after we launched Manistar, I came to the U.S. And I traveled from West Coast to East Coast. And we did more than 15 user meetups in six different cities. Yeah. So, we just want to do that because Manistar is the general AI agent, which means we can do many things. But in the end, even for us, we don't know what the users are using us for. [SPEAKER_01] Right. Yeah. So, we think it's super important to get to know them, to do face-to-face communication. That's the initial intention to open international offices. But later, we found out it's really important to open sites in different countries. Because in different countries, there are different cultures, different ways people do business there. So, opening our office there will make us much easier to talk to local markets. [SPEAKER_01] You were telling me something interesting right before this conversation that when you guys opened these offices, it was very important to you that you kept the product and engineering and research all together. Yeah. [SPEAKER_01] Tell us a little bit more about that. In our company, we think it's very important to put the product guy, the engineering, and also the researchers sitting together. [SPEAKER_02] Yeah. So, I just give you an example. Yeah. Back in our previous office, I and P, our chief scientist, and Bean, which is our full stack guy. Yeah. He's a real full stack guy because he's not just a full stack engineer. He does everything. [SPEAKER_03] Our logo is also designed by him. [SPEAKER_03] Yeah. [SPEAKER_03] Yeah. [SPEAKER_03] So, he's a full stack guy. [SPEAKER_03] Yeah. [SPEAKER_03] So, we sit together. [SPEAKER_03] And every day, we will have many, many small talks. He's like, okay, I saw something very interesting. Can we try that? So, when we sit together, it's very easy to try out some new ideas. But I know there are some teams that will just separate different teams. This is a product part. This is an engineering part. We insist that we should sit together. Yeah. So, right now, it's still me, Bean, and P, we sit together. And every day, we have many small talks. And we believe that those small talks lead to all the innovations we did in the past year. Yeah. Especially probably for a product like Manus, which is so far-reaching and so cross-functional. Yeah. So, you said something interesting about how you structure your engineering team. Most engineering teams are structured by product areas or the customer journey. Yeah. How do you guys decide on who works on what in the engineering team? [SPEAKER_03] Our CEO, Panpan, has his own philosophy about managing an engineering team. Right now, we think we have maybe 40 or 50 engineers, and they are in a very big team. We only have one team, the engineering team. So, whenever we have— 50% team. Yeah. When we have a new project, Panpan will just assign someone to working on this project. But after that, they will come back to the big pool. Yeah. No one is specific, just appointed to a feature or a project. We think it's very important because it will make that much flexible. So, a project comes up, and you're like, who here is available? And people raise their hand, you're like, great, you two go work on this particular project. They're done with the project. They come back into the general pool. Yeah. And then it's the next project. Yeah, that's it. Who maintains the code and the product after they're done with the project? [SPEAKER_03] Oh, yeah, that's a good question. [SPEAKER_03] The way we run our engineering team also somehow benefits from the AI capabilities. Because I think before, why we should let the same engineer keep working on the same project is because we need the context. Yeah. It's really hard to read others' code, right? But right now, with the help of AI, it's really easy. [SPEAKER_01] Yeah. [SPEAKER_01] And then it's the next project. [SPEAKER_01] Yeah, that's it. [SPEAKER_01] Who maintains the code and the product after they're done with the project? Oh, yeah, that's a good question. I think the way we run our engineering team also somehow benefits from the AI capabilities. Because I think before, why we should let the same engineer keep working on the same project is because we need the context. Yeah. It's really hard to read others' code, right? But right now, with the help of AI, it's really easy. If we give an engineer, we think, okay, you should work on this, and there is some existing code. So he can just use the AI to read all the code base and explain how the whole project works. Yeah, so it only maybe took him 10 minutes to understand a very complex project. [SPEAKER_03] But before that, it would take days. It's an incredible unlock because now you have everyone, and also people just by definition of having different projects and coming back, they will have more context anyway. [SPEAKER_01] Yeah. Because they're working on so many different pieces. [SPEAKER_01] So I think that's amazing. I hope that's the future of product development and building products. I run global teams. One of the things that I found the hardest about running global teams, in addition to maintaining context, which I think may end up getting solved, is how you create a consistent culture. How do you think about that? How do you make sure that the Tokyo office and the Singapore office and the San Francisco office feel the same when you show up? What are your tips and tricks for that? Yeah, because for Tokyo and San Francisco, mostly they are marketing. Yeah, the go-to-market team. And it's under my watch. So I might be the right person to answer the question. Yeah. I think it's always about how you lead. Because to do management is not just to manage. It's always about making decisions. So you will let every team member know how the company says things and how the company trades different trade-offs through decisions. So when I made every decision, I will tell the related person about why I made the decision. What's my thought behind the decision? And it's very important to telling employees about that. And the second thing is about how we should have a consistent way we do marketing and the way we tell the users about our message. Yeah. So whatever we do in different countries, we will review just in our weekly meeting about, okay, this is the way we do offline events. This is the message we should tell our users. That's how we align. We align on the message level. Yeah. [SPEAKER_01] I love that because so much of the culture is just, as you said, top down from the leadership. [SPEAKER_01] What are they showing and how are they behaving and acting? [SPEAKER_01] And that really shows a lot. Because everyone will learn. [SPEAKER_01] Exactly. [SPEAKER_01] Yeah, exactly. [SPEAKER_01] Everyone will learn from that. [SPEAKER_01] So you went from zero to 100 million in ARR in just eight months. [SPEAKER_01] Yeah. [SPEAKER_01] How? [SPEAKER_03] Oh, okay. Yeah. I know a lot of people are very curious about that. But we are also curious about that, too. Yeah. Because before last November, which is in the first eight months, we nearly put zero marketing budget into the whole product, which means most of the traffic is organic traffic from everywhere. I can just give you some examples. I think it's back in last June. We were at the top place in the App Store of Egypt. Yeah. And we don't know why. And then we just check, we do some research with Manus. And we found out there is some random guy on Facebook. He writes a post with Arabic language talking about Manus, which made us somehow viral in Egypt. Oh my God. And it makes us at the top place. And we didn't know that before the post. And also in Brazil, we have a huge user base there. And the initial wave is two influencers on YouTube. They made two videos about us. And we went super viral. And we didn't know these two influencers until then. So for us, I think a lot of viral waves just came from word of mouth. Because as you just mentioned, the first time you saw our video, you think, oh, this is something new, right? [SPEAKER_03] And they got amazed. [SPEAKER_03] And they will maybe write something or do a video about us and introduce us to their friends. [SPEAKER_03] So I think in the whole first eight months, it's all about the product itself, about word of mouth. [SPEAKER_03] Yeah. So your growth tactic is just build an insanely good product. Yeah, yeah, yeah. That's your feedback. Yeah, because I know in many teams, they will treat growth as a separate sector in their company. But for us, we think the best growth is the product itself. Yeah, it's super important. But definitely after last year, the management team just had some review about the past year. And we saw there are some important decisions in our product that made us more viral. The first thing is about, maybe 20 days before the launch, we decided to build a feature. We call it the session replay. You remember that? Yeah. It's like every task, whenever it's done, when you share the task to your friends, your friends will see the whole session in replay mode, right? It's like, blah, blah, blah, blah, blah, blah. [SPEAKER_03] So I think that's very important. Because, you know: Very powerful. Yeah, when you share a session to your friends, if your friends only see the final step, they will not get amazed. You know, it's just another AI product, right? But when they see the session in replay mode, it's like, oh, what is going on? The AI is controlling the browser. Yeah. [SPEAKER_03] It's every task. Whenever it's done, when you share the task to your friends, your friends will see the whole session in replay mode, right? [SPEAKER_03] It's blah, blah, blah, blah, blah. [SPEAKER_03] So I think that's very important because it's very powerful. [SPEAKER_03] Yeah, when you share a session to your friends, if your friends only see the final step, they will not get amazed. [SPEAKER_03] It's just another AI product, right? [SPEAKER_03] But when they see the session in replay mode, it's like, oh, what is going on? [SPEAKER_03] The AI is controlling the browser. [SPEAKER_03] The AI is writing code, blah, blah, blah, blah, blah. So I think that feature is really important for our growth. [SPEAKER_01] You also have the build first sign up later feature. [SPEAKER_01] Yeah. [SPEAKER_01] It's obviously using Stripe sandboxes for that. [SPEAKER_01] So we love that. [SPEAKER_01] Has that helped at all with your growth? Oh, yeah. I think that part helped the growth of our v1.5 feature. Because in the version of v1.5, we introduce a new way to build websites. Before that, Mascon only built static webpages. Just static HTMLs. But in v1.5, we can build fully functional websites with the front-end, back-end, databases and native AI capabilities. And for that, many users asking us, can you integrate Stripe into the system? Then they can just run their real business on Manus website. Yeah. But as you know, Stripe has the best developer experiences. But even as a leader, because our ideal customer is a non-technical person, right? [SPEAKER_01] Oh, yeah. It worked. [SPEAKER_03] Yeah. So it's really hard for them to understand what's a token, blah, blah, blah, a lot of things. So we think the Stripe sandbox really helped us to make the whole process much simpler. Even for non-technical users, they can build a functional business, an e-commerce website, without any knowledge about the technical details. Once they're done with the website, they can just translate the Stripe sandbox to the real environment. Yeah, it's an incredible feature. [SPEAKER_01] I'm so excited to see all the businesses that are going to happen because of tools like Manus. [SPEAKER_01] It's really amazing. [SPEAKER_01] One of the really cool things about Stripe, actually, is that we get some insight into all the businesses that build on top of Stripe and how fast they grow. And one of the insights from us is that these AI cohorts just have fundamentally different growth curves. [SPEAKER_01] In this new world, how do you measure your success? [SPEAKER_01] Because you don't have that many benchmarks to go off of. I think for us, we care about two metrics. The first is about ARR. Yeah, it's a traditional metric. But we think it's much more important in this era. Because right now, AI is still very expensive. Yeah. So you can't just make a free product. I worked in the industry for more than 15 years. [SPEAKER_03] So in my early years in this industry, what we're always talking about is don't care about the revenue. Just bring the users in. After you have millions of daily active users, you can find a way to make money, right? But today, I think things have changed. You must ask the users to pay for your product in the first version. Yeah, because the AI is expensive right now. So it's really important to earn money, not just for revenue, but to prove your PMF. Someone will only pay if they feel the value is real. [SPEAKER_03] Yeah, so we think it's super important to focus on the ARR because that means a lot. Not just revenue, but also for your PMF. [SPEAKER_03] The second metric we are tracking is about the user's rating. [SPEAKER_03] Because after every task is done. User ratings? [SPEAKER_03] Yeah, we will ask the user to rate the task one to five stars. [SPEAKER_03] And that tells us a lot about our quality and performance. Because you can always see these ratings through different segments. Yeah, by different features, by different countries, by different pricing tier. [SPEAKER_01] Oh, that's super useful. Yeah, that tells us a lot. So right now we are following these two metrics and we are analyzing every week. Yeah. [SPEAKER_01] So speaking of ARR, how did you think about pricing? Pricing is actually pretty important. But we don't know what's the best approach. We learn from each other. [SPEAKER_03] Actually, if you are looking at today's cutting-edge AI applications from a yearly scope, you will see everyone changes. Yeah. Sometimes we will learn from others. And six months later, we found that they learn from us too. Yeah, everyone just changes their pricing plans. [SPEAKER_01] Everyone's looking at the other. Yeah, yeah, yeah, yeah. So we don't know the best practice here, but we think it's very important that you keep something in your mind. I think in today's AI world, there are some very heavy users, very powerful users. They will feel they get much more value than other users. So they're willing to pay more. So always leave a space there for users to be charged by usage, usage-based billing. It's very important. [SPEAKER_02] Yeah. Even if you have a monthly subscription plan, always leave a space there for users to pay more. [SPEAKER_01] Yeah. [SPEAKER_01] It seems unbounded in some ways. [SPEAKER_01] There are users who are willing to pay a whole lot more. [SPEAKER_01] Yeah, yeah, yeah, yeah. [SPEAKER_01] Which is- And also, I think this is not just about the revenue. It's also about signals. [SPEAKER_03] Because for us, our top plan, our top plan when we launch the pricing plan is $200. [SPEAKER_03] We think it's a lot. [SPEAKER_02] Yeah. [SPEAKER_03] Even if you have a monthly subscription plan, always leave a space there for users to pay more. Yeah. It seems unbounded in some ways. [SPEAKER_01] There are users who are willing to pay a whole lot more. Yeah, yeah, yeah, yeah. Which is- [SPEAKER_03] And also, I think this is not just about the revenue. It's also about signals. Because our top plan, our top plan when we launch the pricing plan is $200. We think it's a lot. After we launch the plan, and we leave an option for our users, which is that you can recharge your account after you hit the limit. And luckily, we found out there are some users who would pay us $5,000 a month. And that's a super positive signal to catch because someone paid $5,000 a month, which means the output of the minus may value 10 times more, right? So we will do user interviews, and we will find out, okay, what are you using us for? Why do you think it's so valuable? And that provides us many insights for the future product roadmap. [SPEAKER_01] Any insights you can share from those conversations? Definitely. Why we made minus 1.5 is because we are looking at users who will build webpages, even static webpages in minus. And they spend a lot. And we ask them, why do you want to build these webpages again? [SPEAKER_03] And they told us in their country, it's very hard to find engineers, even outsourcing engineers, to build a business website for small, medium businesses for them. [SPEAKER_03] So for them, spending $1,000 on matters is super cheap. [SPEAKER_03] Cheap, yeah. So that gave us some inspiration because we live in a very big country. And every day we talk about someone from the U.S. is also full of tech people, right? So we don't feel that pain. [SPEAKER_03] But the pain exists in the real world. There are people we never met, but they have this pain. And after we understand that, we think, okay, maybe we should make our website builder much more powerful. That's why in minus 1.5, we introduce fully functional website building. Yeah. Front and back-end databases. A lot of things. Yeah. [SPEAKER_01] What a fantastic insight. [SPEAKER_01] You're just bringing access to parts of the world that just never had it. And folks can just start their own businesses entirely on top of Minus. Just even five years ago, the bottleneck to being able to execute an idea like this or do a task or be more efficient was coding. Yeah. That's no longer the case anymore. Yeah. [SPEAKER_03] What do you think is the bottleneck now? [SPEAKER_01] Is it people's imagination? [SPEAKER_01] Is it being able to prompt? [SPEAKER_01] Yeah. I think right now, the bottleneck is still the capability of defining a program. Yeah. Because solving a problem is not a problem anymore. But to define a problem is somehow, I think most people are still lacking this capability. Yeah. Because in today's world, the AI still needs you to ask the question, to assign the task to them. So you have to be the one to watch around, what's the pain? What's my customer's pain? And what's our organization need, right now? And you can just generalize questions from this real user's pain or what the real needs inside your organization. [SPEAKER_03] And give the task to the AI, ask it to solve it for you. [SPEAKER_03] But if you can't see these questions or problems in your life, you don't have any task for the AI. [SPEAKER_03] So I think that's somehow still the bottleneck. Do you think eventually AI will also help us ask the questions themselves? Define the problems? [SPEAKER_03] I think maybe. [SPEAKER_03] Yeah, yeah, maybe. [SPEAKER_03] But it is still, I think it's very important for us to learn. I think actually it fundamentally changes the way we learn and we execute on our daily jobs. [SPEAKER_01] Indeed. [SPEAKER_01] Yeah. [SPEAKER_03] It's, I think before it's only a few people can be the manager. [SPEAKER_03] Yeah, because definitely not everyone can be the manager because that won't make the company run it. [SPEAKER_03] You need people to do the work. [SPEAKER_03] Yeah. [SPEAKER_03] But right now, I think everyone can be the manager, right? [SPEAKER_03] Because if you know how to manage, you can manage 10 or 100 agents working for you. [SPEAKER_03] So I think this is a new era for everyone to learn how to be a good manager. [SPEAKER_01] This is maybe a great moment for my final question. [SPEAKER_01] Yeah. [SPEAKER_01] So fast forward 20 years in the future. [SPEAKER_01] AI is doing a lot of the tasks and the execution. [SPEAKER_01] Yeah. [SPEAKER_01] What do you hope the world will look like? [SPEAKER_03] I know, right, a lot of people are afraid of that future because, oh, AI is taking my job. [SPEAKER_03] But if you're looking maybe 100 years ago, when we just invented cars, a lot of people were saying that too, right? [SPEAKER_03] When we invented the machines to lift the heavy, heavy scenes, because, oh, I don't have my job. [SPEAKER_03] Because before that, maybe we need 20 people to lift some very heavy scenes. [SPEAKER_03] And right now with a machine, one man can do that. [SPEAKER_03] But in history, after we have cars, we have machines. [SPEAKER_03] After we somehow release these labor from these tedious work. [SPEAKER_03] There are so many scenes we found out to do, arts, many new scenes, new business models, right? [SPEAKER_03] There are so many scenes. [SPEAKER_03] So I think AI is somehow the same. [SPEAKER_03] I think 20 years later, there are so many tedious work we don't have to pay our attention on. [SPEAKER_03] But definitely there will be more and more important scenes for us to do. [SPEAKER_03] Yeah. [SPEAKER_03] Yeah, I think it's a beautiful thought. [SPEAKER_01] Maybe it'll let us get a little closer to what it means to be really human. [SPEAKER_01] Well, I'm so excited for you and Manis to help us get to that future vision. [SPEAKER_01] Yeah. [SPEAKER_01] Thank you so much, Tao. [SPEAKER_01] This was fantastic. [SPEAKER_01] Yeah. [SPEAKER_03] There are so many scenes. [SPEAKER_03] So I think AI is somehow we're the same. [SPEAKER_03] I think 20 years later, there are so many tedious work we don't have to pay our attention on. [SPEAKER_03] But definitely there will be more and more important scenes for us to do. [SPEAKER_03] Yeah. [SPEAKER_03] Yeah, I think it's a beautiful thought. [SPEAKER_01] Maybe it'll let us get a little closer to what it means to be really human. [SPEAKER_01] Well, I'm so excited for you and Manis to help us get to that future vision. [SPEAKER_01] Yeah. [SPEAKER_01] Thank you so much, Tao. [SPEAKER_01] This was fantastic. [SPEAKER_01] Yeah. [SPEAKER_01] I really enjoyed chatting. [SPEAKER_01] Thank you, Aga. [SPEAKER_03] Me too. [SPEAKER_03] Yeah. [SPEAKER_03] Yeah. [SPEAKER_03] Thank you. Thank you.