Chapter 4 — My Take on “Copycats, Masters, and Makers”
I went into this chapter expecting something technical. After all, Kai-Fu Lee previously laid out the four key drivers of the AI revolution: data, entrepreneurs, AI scientists, and government. Chapters 2 and 3 covered data and entrepreneurship. So now? Time for the scientists — or so I thought.
And yes, that’s how it starts. Lee points out that in the early AI breakthrough phase, you needed a few genius minds, like in the days of the Manhattan Project. One exceptional researcher could be worth an entire research department. But once the big discoveries are made, the dynamic flips: now we need armies of solid engineers who can tinker, test, and improve.
That’s where China shines.
From Copying to Critical Mass
Chinese engineers, for a long time, were seen as implementers, not innovators. And I still think — for the most part — that’s true. They’re world-class at testing and refining what others discover. But that’s not the whole picture anymore.
Kai-Fu Lee brings hard numbers: Chinese researchers now publish nearly half of all AI papers, and a growing number are being invited to top-tier conferences. That tells me this: they’re not just catching up — they’re stepping into the arena.
Still, Lee’s perspective has a bias. I mentioned before — his heart is clearly closer to China, so naturally he highlights their strengths. But if we step back, the U.S. still has an edge in several critical areas:
- Deep-tech startups (OpenAI, Anthropic, etc.)
- Elite research hubs like MIT and Stanford
- Dominance in hardware (NVIDIA, AMD, Intel)
And let’s not forget: it was those American startups that made ordinary people like me realize — “Wait, we’re in a new era now.”
So yes, China is rising. But the balance hasn’t tipped yet.
AI Is Not a Silo
There’s one thing this chapter really drove home: AI can’t be analyzed in isolation.
It’s not just code. It’s also:
- Compute power
- Data access
- Regulatory structures
- Cultural norms
Lee gave the example of a Chinese student who once studied under a streetlight because the dorm cut off electricity at 11 p.m. That same student now has access to:
- arXiv.org and other preprint servers
- Global research updates in real-time
- Forums and tutorials that explain the latest architectures
- Toolkits and libraries shared via open source
But — and this is my own addition — open science isn’t unconditional. It flows when institutions see benefit in openness. But if the U.S. or China stumbles on something truly disruptive, you can bet they’ll lock it down. Science serves ideals — but also serves power.
And yet, the Chinese engineers do more than just consume what’s shared. They often refine, test, and improve it. That’s not trivial. And increasingly, they are contributing their own discoveries — becoming makers, not just followers.
Final Reflection: The Age of Engineers (And Maybe Me)
This chapter reframed AI for me. The era of lone geniuses is giving way to an age of competent, persistent engineers — most of whom will never make headlines. And that’s okay. That’s how real change happens.
And maybe that’s also my own journey. I’m not an AI expert. But I am someone who has followed technology as it transformed through:
- Simple Linux servers
- The abstraction leap of virtualization
- The radical shift to containerization and Kubernetes
- And now — perhaps the most disruptive of all — the age of AI
So maybe I am one of those future AI engineers, trying to understand where I fit. I don’t have final answers. But with every chapter, I ask better questions. And maybe that’s all I can hope for:
To hold a perspective — and let it evolve.