This April Fools Day, I posted the following on social media:
After much serious deliberation, I’ve decided to leave the semiconductor industry.
I’m going to become a fitness coach.
I’ve done the market analysis: AI in the Bay Area is growing fast intellectually, but Bay Area waistlines are growing faster. AI trends toward cost reduction and headcount cuts; caloric intake trends steadily upward with no signs of reversal. The market verdict is clear.
Currently studying for my coaching certification. I’ve already run fitness assessments on a few Bay Area friends — results are not encouraging. Silicon Valley engineers have a body fat percentage that makes chip yield rates look impressive by comparison.
DM me if interested. First ten sign-ups get a discount.
Some friends knew it was April Fools. Others took it seriously. But while interacting with people’s reactions, I found myself thinking about two real questions.
1. Am I actually pursuing what makes me happy?
Writing that joke, I found myself mentally walking through various careers and possible lives.
My honest conclusion: my current life isn’t exactly what I’d want most in some ideal world, but it’s a genuinely good one among the many possible paths. Where do the small dissatisfactions come from?
I think they come from Nietzsche’s camel. Carrying too much, wanting too much, working against myself.
Nietzsche described three metamorphoses of the spirit: the camel, the lion, and the child. The camel is the bearer — carrying obligations, social expectations, the weight of “thou shalt,” trudging silently through the desert. The lion is the rebel — using a powerful self and will to break external constraints, claiming the freedom of “I will.” The child is the creator — forgetting the self, creating new values through play, living in the fullness of the present.
To address the current constraints, the path forward is either becoming the lion or becoming the child.
The lion means using hard-won capability to overpower what’s unsatisfying — strong enough to have genuine choices. It’s the path society expects, and it’s a clear one. The child is a different kind of completeness: giving up the socially sanctioned success metrics, no longer measuring yourself by salary, title, or influence, and instead listening to what your heart and body actually want — slowly arriving at a forgetting of self, finding answers in inner fullness. The child is the harder path, because what you’re fighting isn’t external obstacles but the evaluation system you’ve internalized over years.
I think of Zhang Xue. He started as a motorcycle repair apprentice, and at 19 chased a Hunan TV crew for over 100 kilometers on his motorcycle just to get their attention. He co-founded a motorcycle company in 2017, then in 2024 resigned — from the company he had built himself — because investors wanted quick monetization while he insisted on racing and R&D. He started over from scratch, founded his own brand bearing his name, and in March 2026 won back-to-back races at the World Superbike Championship in Portugal — the first Chinese brand to do so.
Zhang Xue isn’t without ambition. He aimed his ambition at what he genuinely cared about. His “child” isn’t naivety — it’s radical honesty.
I don’t see myself as a lion right now, and I haven’t become the child yet either. My current state is more like a camel who has become aware that it’s in camel mode. I thought finishing my PhD would let me exhale, let me start living at my own pace. Instead I found the camel’s desert had just shifted. The burdens changed; the posture didn’t.
Becoming the child takes courage — not waiting. Not “I’ll let go once I’m strong enough,” but choosing to live honestly before you’re strong enough. I don’t know if I have that courage, but it’s a question I want to take seriously.
2. Has technology (AI) actually made society and humanity happier?
The other question: has AI actually made people happier? I don’t feel like it has.
Everyone marvels at AI’s capabilities and productivity gains, but humans need a sense of accomplishment from their work. AI is taking over more and more tasks — and often the very tasks from which people derive a sense of achievement and control. Efficiency goes up; the feeling of “I did that” gets diluted.
The deeper question is: where is all that AI-generated productivity actually going?
I think about the three genuine scarcities humans face: material scarcity and inaccessible healthcare, structural injustice, and inner poverty of meaning. The most visible AI applications — faster content generation, smarter search, better recommendation algorithms — which of these do they address? Most land in the “nice to have” category, not the “need to have.”
Working in the semiconductor industry makes this more concrete for me: the feedback loop is too long. A chip goes from design to tapeout to mass production to reaching a user’s hands in three to five years. It’s hard to trace where my work actually ends up, whether it changed anything real for anyone. That distance is a slow leak of meaning.
This isn’t unique to semiconductors — it’s a property of large technical systems: the more complex the system, the longer the chain between an individual and their impact, the harder it is to feel that your work matters to actual human lives.
This moment has happened before, more than once.
The Industrial Revolution’s machines promised to liberate people from hard physical labor. In practice, workers moved into factories, worked longer hours in worse conditions, and the efficiency gains flowed to factory owners. What workers got was what Marx called alienation — separation from their labor, from its products, from each other.
Early twentieth century home appliances — washing machines, refrigerators — were supposed to free housewives and give them leisure time. Sociologist Joann Vanek found that cleanliness standards rose alongside the appliances, and women’s time spent on housework didn’t actually decrease.
Social media in the 2010s promised to connect people. Instead, loneliness rose — the “loneliness epidemic” is a term that emerged in that decade. The quantity of connections increased; the quality declined.
The pattern is consistent: technology raises speed and access, but simultaneously raises expectations and complexity. Happiness doesn’t increase proportionally — sometimes it declines. “Faster” doesn’t equal “better.” “More” doesn’t equal “more satisfied.”
I don’t know where the answers are, but I think this question — whether technology is actually making human lives more meaningful — is worth asking more seriously than “how capable is the AI?”
It’s interesting that these two questions turn out to be two sides of the same question. What I worry AI is taking from workers — the sense of achievement, of agency, of connection to what you make — is precisely what I feel myself losing in my own life. The camel’s problem and the AI’s problem have the same root: efficiency is rising, but where is the human?