“This post was originally published on my personal blog at Nookix Blogs. This is a repost.”
The story is a bit long, so if you don’t have time, feel free to skip to the actionable advice section at the end. But I believe my real story will help you in some way — at the very least, it can save you from taking some detours.
From Entrepreneurship to Corporate Career to Layoff
I started my own business before finishing graduate school and secured 1.2 million dollars in angel investment. Given my family background and the fact that I was still a student, 1.2 million was no small sum to me. But I messed up. Not only did I mess up, I also dragged my roommates and college classmates into it — they joined me because I encouraged them. We worked on a smart home project. Maybe the idea was too ahead of its time, or maybe I just severely lacked experience. Either way, the project inevitably died after barely hanging on for two years. Fortunately, none of us ended up in debt. After that, we all went back to regular jobs, got married, had kids, and lived ordinary lives.
I was no exception. I swallowed my pride, calmed down, and became a product manager. Along the way, I moved between startups and public companies, rising to senior product manager, project lead, and department head. Most of it went pretty smoothly. In 2021, six years into my career, I had just gotten married, was carrying a heavy mortgage, and had bought my first car. My income was peaking.
But that downstream trend was interrupted in 2022. That year, I felt deeply that the internet industry was hitting its ceiling. Layoffs and “cost reduction and efficiency increase” were the inevitable trends. Yet I needed money — I needed to save for the arrival of my child, and my financial foundation was still shaky. There was a clear path ahead: choose a company with more growth potential in a sunrise industry and continue chasing promotions and raises. So I jumped to a pre-IPO company and managed to get a decent chunk of options and stock. Later, the company did go public. At its peak, my stock was worth a few million dollars— not a fortune, but like everyone else, I started getting a bit inflated. I began looking at bigger apartments in better neighborhoods.
Fortunately — or unfortunately — later on, our business unit was too cash-intensive, and the boss decided to shut it down. So I landed on the layoff list. This was my first layoff experience; before that, I had always jumped ship voluntarily for better opportunities. The layoff wasn’t just a mental blow — my shares were still locked up and I couldn’t exercise them. According to the agreement, they had to be taken back. Illusory wealth, gone in a puff of smoke. Thankfully, I hadn’t bought that bigger house, otherwise my mortgage burden would be even heavier. But by then, the job market had become even worse after the COVID pandemic — layoffs and unemployment were everywhere, and it was horrifying.
I have to admit: as long as the bubble of chasing promotions and raises hasn’t burst, I’ll always have unrealistic fantasies about it. Even though there’s always been a voice deep inside telling me to build my own business and pursue my own dreams, I kept suppressing it. After all, the environment is so bad right now; starting a business rashly isn’t a wise choice. Poetry and the distant horizon can’t always put food on the table. As a husband and a father, you have to deal with daily necessities. You have to be conservative and prioritize survival.
Awakening and Replanning After the Layoff
After being laid off, I got some severance and took my first gap month ever — just rested. But this involuntary departure was a wake-up call. The workplace is a brutal arena. Your loyalty and hard work can be rewarded by being kicked to the curb at any moment. After you’re kicked out, all your investment instantly goes to zero. You no longer have any connection to that company, and it’s as if you’ve been reset overnight. Not everyone is a child of destiny. Not everyone will be lucky. Ordinary people without assets, who rely on working for others and chasing promotions, will find it very hard to achieve financial freedom. That path may well be a dead end. Some jobs just disappear as you’re working them, as companies prefer cheaper young people or even AI. And financial freedom remains a distant dream.
Everyone knows these truths. But actually breaking that obsession and resolutely embracing your own understanding is another matter entirely. Unless pushed to the edge, most people are unwilling to voluntarily leave their comfort zone. After this layoff, I became completely clear-headed: I have to find my own business and completely break away from working for others.
But because I had been fully committed and focused on climbing the corporate ladder, I had never seriously tried any side hustles. Suddenly, I didn’t know where to start. During that gap month, I tried a few side projects — like a TikTok account and an investment blog — but never had stable income, and I couldn’t support my family’s expenses.
After discussing with my wife and friends, I decided to find another job first as a transitional step, working while searching for direction.
So I landed a job relatively smoothly, albeit with a pay cut (after COVID, the job market worsened, and salaries are now negotiated down). But I didn’t take it too seriously. Over the next two years in my spare time, I started tinkering with side hustles. I analyzed my situation: I’m an internet product manager, I’m busy with my day job, I don’t have a lot of free time, and although I’m in the tech circle, I don’t know how to code. So I was better suited for lightweight projects.
Side Hustle Exploration and Vibe Coding
At the end of 2024, I launched my first side project, VeryCareer. I used a third-party LMS to build a course website, and also hired someone to develop a conversational, step-by-step plugin. I sold an innovative Excel skills course that emphasized learning by doing. The reason I chose this was simple: at my previous company, I had led a similar online education project, so I knew its operational model and business ceiling pretty well. But when I actually started doing it as a side hustle, I realized that entrepreneurship requires a jack-of-all-trades, not just a specialist in one area. Producing the course content myself was completely different from giving orders to a content subordinate when I was at a company. This project was painful for me — even though I understood the business model of online courses very well, content creation was entirely outside my circle of competence. I hired some help, but the cost and results didn’t meet my expectations for an MVP. Plus, ChatGPT-3 burst onto the scene around then, and I quickly realized that selling courses was becoming obsolete. I spent tens of thousands on ads, got terrible data, and shut it down soon after.
Maybe the timing was just right — the AI spring came faster than expected. After that, various AI models emerged, along with IDE tools like Cursor and Anti-Gravity. In May 2025, me, someone who knew absolutely nothing about code or basic software concepts (I didn’t even know how to resolve a domain name or use git), started learning by doing with the help of the not-so-intelligent Gemini 2.0 and Cursor. I began my second project — an audiobook website called Nookix. The idea was to use AI to break down a large number of books and user reviews, then generate high-quality 30-minute audio podcasts that thoroughly explain a book to users.
I have to admit, I was pretty proud of myself at first. After three months of staying up late, I finally launched a website from scratch with zero foundation. But as everyone knows, early AI models had terrible hallucinations and weak coding abilities. Plus, with no experience, my early technical architecture was a mess, and I kept changing things. Eventually, I saw from Google Search Console’s crawl errors that my code was a pile of garbage. After months of work, traffic was pitiful. I was about to shut down the project.
Even worse, for someone who doesn’t understand technology, building a content production pipeline for book audio was no easy task. You also have to do a lot of cleaning, a lot of review, proofreading, and revision. I was pulling my hair out. It was too hard. I didn’t know how to solve many problems. So I put the project aside, even considering shutting it down.
Then came the end of 2025. With the release of Gemini 3.0 and the launch of Anti-Gravity (which came with Gemini 3.0 integrated), I was thrilled. Problems that I used to struggle with in the Gemini 2.0 era — talking to AI in Cursor for ages without a fix (you know, that back-and-forth, fix-one-thing-break-another frustration that drives you crazy) — Gemini 3.0 solved almost instantly. Fast and good.
That made me realize the timing was gradually maturing. I have to admit, I got a bit overconfident about my technical abilities. I thought this era belonged to product managers — that I could quickly turn my ideas into products. I had so many ideas I wanted to realize, especially those product dreams I’d always wanted but never had the skills to build.
The Expanding Bubble of Projects and Capability Illusions
So over the next three months, I launched three products in a row: Stockxy, an investment tracking tool for value investors; TikClass, an interactive, conversational classroom SaaS tool for B2B customers (yes, I just can’t let it go — in my previous online education project, I had a friend develop an interactive classroom plugin for me, and I rebuilt it using AI on that foundation); and Domainay, an AI domain name generation, registration, and product launch platform.
Seriously, AI coding gives you immense instant gratification. The ultra-fast feedback, the what-you-see-is-what-you-get — it makes you unable to stop. Honestly, after so many years of working, I’ve rarely had the chance to experience such clear, fast feedback (most jobs have slow feedback loops, and product management especially). This rapid feedback massively stimulates dopamine and becomes addictive. I actually forgot for a while that I was doing side hustles to make money, not to satisfy my damn desire for exploration.
Become a Medium member
I gradually realized that a good idea and a successful project are separated by a lot. A whole lot. First, you have to develop a stable, usable product that can withstand user testing. Second, if you’re offering a content-based product, you also have to solve content production. Third, you have to figure out marketing and operations. If you rely on SEO, you have to understand all the ins and outs of SEO. If you rely on paid ads, you need some understanding of advertising and be able to control budget and ROI. If you rely on social media, you have to carve out time to run accounts consistently and get enough traffic.
Everything piles up, and complexity explodes geometrically. Doing each thing well requires enough time. If you have a weakness in any area, it’s hard for the project to succeed. And after ten years of corporate training, I’ve turned myself into a highly professionalized person. What I’m really good at — and can actually do myself — is only product definition and product design. And without the various scaffolding provided by a company, many things become harder.
Focus, Return to User Value
So after three months of tinkering without much to show, I called cut myself. I paused TikClass, Stockxy, and Domainay (they’re still online, but I’m not spending time maintaining them; I’m open to selling them). First, for a pure tool-based B2B product like TikClass, sales might be more important, and I’m not great at that. For Stockxy, I may not be familiar enough with the stock investment industry to build a good tool — it would require a lot of coding, which isn’t my strength. And for Domainay, the competition is too fierce; products like that need to win through SEO, which takes a long time to build.
I focused all my energy on one project: Nookix. Because I believe that no matter how AI develops, reading is an eternal need. Only the way we read will change with technology. And given my past work experience, I have my own understanding of book and content products. I think I can go deep in this field. What I need is to pour all my time and passion into this project, saturating it with my effort.
So I spent nearly a month completely refactoring Nookix, identifying and fixing all kinds of hidden technical issues (sure, there might still be some hidden problems, but the experience is now very smooth). After going through those other projects, AI coding is no longer a complex thing for me. But precisely because I’ve gained rich project experience, I have to say this: as an indie developer, you must stay humble about vibe coding. First, more isn’t always better — too much can easily cause you to lose focus. Second, faster isn’t always better — fast is fun in the moment, but you’ll pay back many debts. Another important point: after stepping on so many landmines, I’ve learned a lesson. If you yourself are an idiot or an outsider, don’t expect to hire an expert and have your business take off. You have to become the expert yourself, and then use AI or hire an expert to help you execute that task accurately and efficiently.
For example, the copy on Nookix’s website — even though AI gave me a draft, I ended up going through it word by word, incorporating keywords, deeply researching competitors, user psychology, and user value. I used my ten years of product, operations, and marketing experience to harshly review AI’s output, just like I used to manage my subordinates. Given the current maturity of technology, the best models might achieve 90 points, leaving 10% error and uncertainty. But you have to realize that the business world values certainty and reliability (that’s why fully autonomous driving is still a long way from mass adoption). To make up that remaining 10%, you must be the final gatekeeper.
So those ordinary people empowered by AI, who seem to have massively expanded their capabilities, are actually still far from real business practice. The business world doesn’t want fast, massive quantities of 60-point work. It wants a small number of stable, reliable 90-point efforts. In the end, those mediocre 60-point efforts will be ruthlessly eliminated, just like cheap fast-food restaurants being eliminated in an era where consumer health consciousness is rising.
So when the recent OpenClaw hype came around, I didn’t blindly jump on the bandwagon. Because I realized that if you have an important, tedious task to solve, the best practice isn’t to communicate with AI through vague language. It’s to turn your idea into a Python script first, then embed a large language model’s API into that script, letting the model precisely solve specific problems while using engineering methods for everything else. The results will be far more stable and controllable than OpenClaw, and the cost may be ridiculously low. So when my friend told me he wanted to play with OpenClaw but didn’t know what to use it for, I told him that what he needs isn’t an OpenClaw, but a specific, clearly defined, valuable task. There’s no need to treat everything like a nail just because you have a hammer.
On Working for Others vs. Entrepreneurship
For me, this year and a half has been a rehearsal for entrepreneurship. I no longer think about when to start my own business. Instead, I assume that I’m already doing it, right now, and I get to work.
It’s been ten years since I last started a business, and I almost forgot what it felt like. But I have to say, the feeling is completely different from being an employee. Almost every day brings new problems that you can’t avoid. You have to explore effective methods within a week or two, learn the tools, and get the process working. For example, in a previous project, I needed to clean tens of gigabytes of data. At first, I didn’t even know how to open a database file. Then I started tinkering with DBeaver, using SQL to join tables and clean dirty data, learning as I went. Sometimes I was so busy I forgot to eat, sitting for hours at a time. My wife always comforts me: “Each difficulty is tough, but you get through each one.” Luckily, for many problems, if I’m truly stuck, I put them aside for a while. Then suddenly, I find a new method or a completely new approach, and the problem solves itself.
I once watched an interview with a well-known entrepreneur. The reporter asked him for his secret to starting a business. He said, as an entrepreneur, you shouldn’t wait until you’re ready. The most important thing is to learn quickly and solve problems quickly. I deeply understand that philosophy now. Because only after you jump in do you realize that every day brings endless new problems waiting to be solved, many of which you’ve never encountered before. You have to learn to cope and solve them fast. That’s the biggest difference from being an employee: a job pursues certainty, while entrepreneurship faces massive uncertainty. If you can’t accept that uncertainty, I advise you not to start a business or envy so-called one-person startups or indie developers. If you’ve recognized this issue, then from day one, you should set out on the road and consistently have the determination to solve problems as they arise. As Warren Buffett said, whatever you want to do when you’re financially free, you should start doing it now.
Now, I’m about to “celebrate” another layoff. Yes, again. HR has already talked to me, and the department I led has been disbanded. I’m taking it calmly. I feel that this time, I have no attachment left to the corporate world. I also know deeply that this may be a turning point in my life. I’ll be heading down a somewhat different path, with a clearer life goal. Even though I’m not fully prepared, I’ll explore as I go.
Final Thoughts and Some Lessons
Before I knew it, a year and a half had passed since I started tinkering with side hustles. Some might ask, “How come none of your projects worked out?” What I want to say is, this is the real reality of entrepreneurship. You have to accept failure — it’s the norm. Only through constant trial and error will you learn which projects suit you and which don’t, and gradually define your circle of competence and your boundaries.
This year and a half has been intensely demanding. Honestly, the speed at which I’ve learned new things is probably five times faster than before. I’ll say this: learning really comes best from doing. That’s the fastest way. We should be grateful that in the AI era, every individual has equal learning opportunities. As long as you’re willing to climb, the pace of learning and improvement is astonishing. AI will also fully amplify the energy of those who love to learn.
Quick plug: if you’re interested in Nookix, you can search for it online — it’s the first result on Google. There are many book lists, many suitable for entrepreneurs to listen to. You’ll find plenty of sharp, insightful comments, way more interesting than dryly listening to someone recite a book.
Here are some lessons about entrepreneurship, side hustles, and AI:
If you have an entrepreneurial heart, don’t wait. Start now by putting yourself in that mindset.
Choosing a side project is like driving a car. You’ll know which car suits you and which doesn’t after one lap — but don’t pick the hardest problem. Entrepreneurship isn’t about proving how awesome you are.
It’s very hard to achieve financial freedom by working for someone else, because you don’t own assets — unless you get extremely lucky.
If you lock onto a project, concentrate all your energy and go all in. You don’t need ten mildly unsuccessful small projects.
Learning, doing, and specializing at the same time is the norm in entrepreneurship. You must embrace it.AI’s 60-point products will be brutally eliminated. The more experienced you are, the more you should be reserved about AI output. It can greatly improve efficiency, but the human must be the final gatekeeper.
AI is an opportunity (a leverage effect) for ordinary people to start businesses. But remember, the most important thing in entrepreneurship has never been technology — it’s user needs.
Top comments (0)