Before I knew it, the Spring Festival holiday had reached its final day. I had already started this 2025 review earlier — if I didn’t finish it now, it would probably end up abandoned in drafts forever. So I decided to stop procrastinating, filled the food and water bowls for the two cats at home, and went back to my study to collect my thoughts.

For many people in this industry, I imagine one word defined 2025: AI.
But when it comes to writing — a form of self-expression — I still resist letting AI do it for me. Even if, in the near future, AI can produce writing without that recognizable “AI tone,” the intention matters. Writing, for me, isn’t about efficiency.

That said, this post was co-created with AI in a different way. I asked ChatGPT to generate six interview questions based on our entire conversation history. I answered them myself, then fed those answers back into the model to form new memories. None of these words have been polished by AI. If you read closely, you can probably picture a slightly unkempt, unshaven middle-aged man — hair overdue for a haircut, holed up at home all Spring Festival — typing under a desk lamp.

Model: ChatGPT 5.2

---

1. Identity and Inner Change

Where did you truly grow this year, beyond just being busier or more skilled?

Starting with a question like that is almost unfair.

If this were a job interview, it would actually be easy to answer. You look at the JD for a more senior role, extract the skills, turn them into a framework, and you’re done. In the workplace, skill improvement is growth — especially soft skills. For example, learning how to stay calm in a requirements meeting instead of slamming the table and calling it “big-company syndrome.”

But in life, real growth is harder to pinpoint. I’ve lived inside my own habits for years. Maybe things changed, but I didn’t notice.

So I did something silly: I retook an MBTI test. Surprisingly, I’d shifted from ESFJ-A in past years to INFJ-T. That alone was interesting.

I asked AI to analyze why. Thanks to the fact that I’ve been using ChatGPT since 2023 and all our conversations are preserved, it could track how my questions evolved. It gave me this summary:

Over the past year and a half, I shifted from a collaborative execution role inside a company to an independent product builder. That didn’t just change what I worked on — it changed how I thought and where I drew energy from. Designing products, refining architectures, writing reflections, and asking what kind of builder I want to be made solitude productive instead of draining. My focus moved from optimizing existing systems to imagining tools that didn’t yet exist. Naturally, my attention drifted from immediate details to patterns, motivations, and long-term direction.

Reading that, two thoughts struck me:

• I’m even more convinced that the moat for AI applications lies in memory and data.
• It’s incredibly hard to evaluate your own growth when you’re inside it.

Coming back to the original question, I think this shift itself counts as growth. Not because one personality expression is better than another, but because it made me clearer about what I actually care about — and more deliberate about choosing my future. Knowing what to do, and what not to do.

---

2. Decisions

What was the most “true to yourself” decision you made this year?

Without a doubt: moving from Guangzhou back to Shanghai, going from independent builder to employee again, and shifting from Product Manager to Product Engineer.

When I left Shanghai for Guangzhou in 2024 to pursue AI product ideas, I set aside a decent amount of money to experiment. But I had to admit that making money through independent development isn’t my strength. Independence is easy. Building isn’t that hard either — model capabilities are evolving incredibly fast. The hard part is making something people will pay for. The even harder part is doing that with investors’ money.

Once I recognized that, I cut my losses quickly. I extracted the core module from my product and sold it off — looking back, it was basically just a bundle of skills. Then I started sending resumes again, carrying the invisible label of a “35-year-old failed founder.”

Among offers from Guangzhou, Shenzhen, and Shanghai, I chose to return to Shanghai — to Tesla.

There was definitely some risk in that choice. The offers in Guangzhou and Shenzhen were also AI-related, also foreign companies, and came with hybrid work. They would have been less intense, with a lower cost of living and easier housing. But after nearly a decade at my previous company, everywhere else felt less intense anyway.

Choosing Shanghai again, choosing Tesla, ultimately came down to wanting to play a small role in a larger story.

When I watch Musk interviews or look at what Tesla and SpaceX have accomplished, I still get goosebumps. I know there’s plenty of marketing baked into that narrative. I also know the world is often a chaotic stage run by amateurs. But I’m still drawn to that vision. That’s just part of who I am.

---

3. Values and Direction

What lost importance in your life this year?
Has your definition of success or a good life changed?

In 2025, I completely lost my attachment to other people’s evaluations. Even after returning to a corporate role, I found myself working according to standards I set for myself. (To be slightly shameless, those standards are pretty high — maybe not state-of-the-art, but definitely first-tier.)

Of course, there are still many things in life and work you have to do. But my mindset shifted from:

“I have no choice, so I must not mess this up”

to

“I’m choosing to do this, and I decide how well it’s done.”

Looking inward is the only way to find real inner peace.

As for success or a good life, I find it hard to define a specific state that would qualify. Maybe it’s easier to ask: looking back on this year, am I dissatisfied with where I am? At this moment, honestly, I’m not. Things feel… okay.

The biggest shift, perhaps, is that success is no longer about being recognized. It’s about being aligned with myself.
(Yes, I know that “not this, but that” structure sounds like something AI would generate — but that sentence really did come from me.)

That said, I do worry about how long this feeling can last. Society evaluates success at snapshots in time, but life doesn’t pause at the highlight reel. If I assume an average lifespan of 75, I still have 40 years ahead. Will the criteria change?

Probably. The only constant is change.

---

4. Confusion, Doubt, and Reality

Did you experience a period of confusion this year? What were you questioning?
What moment made you most proud of your choices?

Sometimes I feel like ChatGPT is cheating. It has the entire conversation history and knows exactly where my doubts were, yet it still asks me to revisit them.

The beginning of 2025 was the most confusing period. I’d burned through a lot of money, tried several product directions, and still hadn’t found product-market fit. That kind of hit is different from anything in a job. At work, you might get pressured by bosses or clients, but you still collect a paycheck. Entrepreneurship is different. Every action is your responsibility, and the revenue numbers don’t lie.

They tell you plainly: you’re just not good enough.

I used to think I had strong execution skills. But when tested in the arena of entrepreneurship, I realized the gap was larger than I thought. I knew I should validate demand before building, yet with AI at my fingertips I rushed into features. I knew I should copy proven business models with pixel-level precision, yet pride made me hesitate, like Kong Yiji clinging to his scholar’s robe. The result was imitation without effectiveness.

So many things I “knew,” yet couldn’t turn into results. That’s an execution problem.

Would the outcome have changed if I’d actually done what I knew? Maybe. But that kind of counterfactual thinking only fuels regret. It’s better to ask how to avoid repeating it.

As for what I’m proud of: even after deciding not to rely on entrepreneurship for income, I kept building. I currently have seven projects, all done after work in my spare time. Maybe one day I’ll write separate posts about them.

Coming home each night and building something in code still feels like constructing a city skyline. I’m willing to “waste” my time on things I genuinely love.

---

5. Theme of the Year

If this year had three keywords describing your inner state, what would they be?

• Anxiety – investing time and resources without getting results, and sometimes not even seeing a path forward
• Relief – returning to the job market after two years of startup attempts at 35 and still landing solid offers
• Hope – listening to Musk’s big-picture narratives and feeling optimistic about humanity’s future… and deciding I should exercise more so I can stick around long enough to witness it

---

6. A Note to My Future Self

What is the most important message you want to leave for your 2026 self?

Just one sentence: it’s my zodiac year — keep your head down. Stay steady. Stay safe.

---

Toward the end this piece may feel a bit uneven. It’s been a long time since I last wrote something this long, and most of the writing I do at work these days is AI-generated. But I still think it’s important to keep a small patch of land just for myself — a place to let my mind wander. These unpolished words are the real version of me.

Even if one day we all become digital lifeforms, that won’t matter as much as this: right now, I’m expressing myself.

And with that — wishing everyone a smooth start to the year.