Ten years, and my answer was yes

Sep 20, 2026

Tonight is an evening without a hall and without stage lights. When I was still in Vietnam I used to run offline events so that people could meet face to face, talk to each other, even hug and shake hands. Now I am far away, not in Vietnam at all, so we meet on a Zoom call instead.

I am sitting in San Francisco. Standing, to be accurate. This is the place where, more than a hundred years ago, people crossed an ocean to come here for gold. There is a rush like that again now, and this time it is called AI.

The technical barrier has been falling away #

People used to talk about a barrier, the technical one. It was a barrier almost nobody could step over. You might need to study at university, you had to work at real technology companies, you had to spend years practising before any of it became real to you.

Today you only need an idea, and the nerve to start. Wherever you are sitting on this earth, you need a computer and an internet connection and you can begin. That technical barrier has been falling away.

So the hard part now is not the technology any more. It sits somewhere else. It sits in how we frame the problem, how we ask the question, and how we decide what we are going to build and who it is for.

Two years, more than seven billion đồng in salaries #

I know business well. I started my first company at twenty-one, so by twenty-one I had already founded something. I understand marketing and I understand selling, and I have had some success in the field I have followed for the past ten years, which is affiliate marketing.

From 2024 I carried an ambition that had been with me since I was very young. I read books, I read the news, I followed what people wrote, and I loved technology. I used to wish that one day I would start a technology company of my own.

So in 2024 I founded my first technology company. I had been a business person, a marketer, a salesperson, a community and content person, and from there I stepped into technology and started a company in it. Then I started a second one, both of them in the field I actually understood, affiliate marketing.

Over those two years I worked with engineering teams. I hired two of them to get started. And in those two years I burned more than seven billion đồng on salaries alone.

Working inside my own technology company taught me how much I did not know. I was starting from the beginning. I did not know the product design process. I did not have the language to talk to the engineers on my own team. There were many times I wanted to describe what I had in mind and had no technical vocabulary to say it with, and many times they spoke to me and I did not understand, so I had to look words up and learn things I had never learned before.

Honestly, when I started I did not know where to begin. I only knew I should be doing something. And I believed technology would let me solve a larger problem, at a larger scale, so that what I made would be useful to more people and could serve them anywhere in the world.

There were stretches where I felt completely stuck. I did not know what to do next. I had no advisor and nobody walking me through it step by step. Starting a company is brutally hard. Compared with affiliate work, or anything else I had done before, building a technology company was a hundred times harder.

There were months when I spent more than five thousand dollars just to buy and try AI tools and software, so I could take them apart, learn from them, and see how they had designed the product inside.

The truth is that the money went out every month and none of it came back. I paid salaries at the start of the month and the month was gone. There was always work, I was always busy, and there was nothing to show for it. The launch date never arrived. I was carrying a lot inside me, and the product still had not reached a single end user. The budget, meanwhile, had run out.

Starting again, on firmer ground #

Because of all that, I decided I had to begin again, on firmer ground.

That was when I went abroad. The first subject I chose was English, because I needed to be able to talk to the rest of the world. Then I decided to learn one more language, a programming language.

I learned to program from the beginning, from the first lines of code, from CSS and HTML, until a website came out of it. I learned about AI and machine learning, and I flew to big cities to learn about technology, startups and investing. I spent a few months in Singapore. I spent a few months in London. And now I live in San Francisco, the heart of Silicon Valley, where the companies that made the thing in your hand, or the thing you open every day, come from. Even the tool we are talking through tonight came from here.

The way I learn is to work on real projects and solve problems that actually exist. I talk to users, I talk to my customers, I talk to my community, and I run workshops to stay close to the people I am making things for. I also spend a lot of time with software developers and AI engineers who have the experience I do not, in every city I have landed in.

Record it, then ask until I understand #

In my first days abroad my English was very poor. Even today I still have to work on it daily, and speaking confidently with other people still does not come naturally to me. As for programming, I understood none of it.

I will be honest with you. When I went to events here and listened to software engineers explain how they build their products, how they design their structures, how they work with algorithms, how they build and train AI agents, all of it was new. I had no background in any of it. But the way I am wired is to throw myself in first and find a way afterwards.

In my first year in San Francisco, and I have been here almost two years now, I went to every event I possibly could. More than seventy of them in person, meeting and listening to a lot of very good people.

There were also moments when I felt completely out of place. I felt I was being left behind. There was nobody to tell that to. I do not usually talk about the pessimistic moments, the hard ones. But those were real feelings and I went through them.

At those events, I did not understand the English and I did not know the subject either. Some people listen once, understand it immediately, join the conversation, ask their question, and leave with an action plan by the end of the session. I understood maybe five or ten percent.

So here is what I did. I recorded the whole talk. Late that night I turned the recording into a transcript. Then I took it to an AI and asked, and the way I ask is deep and detailed. I did this over and over. I asked until I understood what the speaker had meant, until I understood the problem being described, and then I asked how to apply it to my own work the very next day.

I knew that if I applied it, if I actually used it, I would learn fast. So every day I improved my own product, to make it sound and make it right. I was not a technical person to begin with. I was not a programmer and I was not an engineer. But I am working towards being a real engineer, and I am stubborn about it, because an engineer is someone who solves problems, helps people, and makes things that are genuinely useful.

Ten years, and my answer was yes #

I once watched a video of Sam Altman, who runs OpenAI. He said that if you decide to start a company, a technology company, you have to be ready for a ten-year journey.

That one landed hard on me. After I heard it I looked at myself. I am past thirty now. Do I have the persistence, do I have the nerve, to give the next ten years to this?

Ten straight years of pushing. What does ten years actually mean? Ten years of weekends that are not weekends. Ten years of working through the holidays. Ten years where some days run from six in the morning until one or two the next morning.

And my answer was yes. I will give ten years to this next stretch of building. I decided to keep going.

Why Build to Own exists #

The first cohort of Build to Own landed in the one month I had off for the summer. I am studying computer science, which is another thing I am doing from the beginning. I decided that in that month off I had to make something that actually meant something.

So I ran Build to Own, to share everything I had learned, everything I had got wrong, everything I had fallen over and then made sense of. While I was starting my own companies I used to wish somebody would walk me through it step by step, so I turned it into eight practical sessions.

And it did mean something. To me first, honestly. And then, I think, to somebody among the hundred-plus people who came into the programme.

Over those eight sessions, more than a hundred people walked in and went through it one week at a time. First they found a problem. Then they set up the working environment of a real engineer, and a lot of them told me that alone, in sessions one and two, was worth it. Then how to size a market, how to research competitors broadly, globally, without missing anyone, and how to find the gap left open in the market they are working in.

Then the first mini apps. Several of them came back to show me three, four, five mini apps they had built, tools that solved the problems they run into every day and the problems their own teams run into. Some of them went and collected those problems from their teams, asking what keeps breaking in the daily workflow, and then built tools that took the repetitive manual work away.

They stood up versions that ran, the ones we call an MVP. They learned to set up automatic payments, how to take money from anywhere in the world, and how to wire in an affiliate programme, so partners can be tracked and paid.

And they learned new ways to distribute. Session seven alone was more than fifty slides. On distribution I did not repeat a single method I have been teaching for the past ten years. I only covered the newest AI-native ways of getting a product in front of people.

These are the things I do myself, every day. That is the truth of it. When it came to setting up a working environment, you saw my screen and I hid nothing. You saw exactly how I run a team of AI agents around the clock.

What that actually costs #

Here is the other side of it. Before this, there was almost no time to rest. I would check what the first AI had done and the second one had already answered, and then the third, and the loop ran from morning to night.

There were days with no time to do laundry, no time to exercise, no time to go out for coffee. It has been a long time since going out for a coffee was a normal part of my day, the way it was when I lived in Vietnam. People think that moving abroad means an easier life. What it really means is that I am very focused.

Four weeks, and ten products #

Eight sessions is four weeks. That is not a long time to build a product. Out there, building an MVP that people can actually try takes a team and a real budget, and often at least three months, usually more.

I have done it that way. The first month goes to assembling the team and hiring. From the second month you start talking about the idea, and the team builds from the second month to the fourth, and only then does the product launch.

Here it was four weeks, while learning, practising and doing the exercises, and while holding down busy jobs on the outside. And they still shipped something and brought it to demo. That is remarkable.

Tonight is run by a very small team. That is the truth. Most of them are doing it after hours, some of them as volunteers, and nobody is being paid. Without them this evening would not run at all, and while they were testing the setup it was two or three in the morning where I live.

Tonight we are not talking about theory and we are not covering material. Tonight belongs to the ten products that made the final round out of more than a hundred people. They will tell you what problem their product solves and who it is for, and you will watch it run for real.