"AI Superpowers" by Kai-Fu Lee – What It Got Right (And Wrong)
Kai-Fu Lee’s AI Superpowers made waves with bold predictions about AI’s rise in China and the U.S. But in 2025, how accurate were they? Let’s break down what the book nailed — and what it missed.
A Tech Visionary’s Wake-Up Call
When AI Superpowers was released in 2018, it offered a bold, sometimes alarming view of the future of artificial intelligence.
Written by Kai-Fu Lee — former president of Google China and one of the most respected voices in AI — the book predicted a fierce tech rivalry between China and the U.S., along with massive disruption to jobs and industries.
But now it’s 2025 — and generative AI has reshaped how we work, create, and live.
So… how did the book hold up?
Let’s explore what Kai-Fu Lee got right, what feels outdated, and how the book still matters today.
What AI Superpowers Got Right:
1. China’s Rapid Rise in Applied AI
Lee predicted that China would dominate “application-first” AI — and he was right.
From AI-powered:
-
facial recognition
-
e-commerce personalization
-
instant translations
-
digital payments
China leads in scaling real-world AI applications, while the West (especially Silicon Valley) has led in research and foundational models like GPT and Claude.
2. The U.S. vs China AI Arms Race
The tech rivalry is still on — and more intense than ever.
The U.S. has led in large language models (OpenAI, Anthropic, Google DeepMind), while China has advanced with companies like Baidu, Alibaba, and SenseTime — though often behind closed walls.
Lee was right: the future of AI isn’t just about algorithms — it’s about global power.
3. Massive Job Displacement in Routine Roles
He predicted AI would hit:
-
customer service
-
transportation
-
simple white-collar jobs
In 2025, we’re already seeing:
-
AI writers, agents, and coders
-
AI in legal research, tutoring, and HR tasks
-
Widespread job reshuffling, not just automation
His concern for economic inequality and workforce disruption was well-founded — and now it’s urgent.
What AI Superpowers Got Wrong (or Underestimated)
1. Speed of Generative AI
Lee didn’t fully predict how quickly large language models (like GPT-4 and beyond) would reshape everything — writing, programming, design, education.
He focused on narrow AI — not the creative, human-like generation capabilities we see now.
While he understood deep learning, he underestimated how “human” AI could feel by 2025.
2. Open-Source & Decentralization
The book assumed major players would be corporate or state-driven.
But in 2025, open-source AI models (like Mistral, LLaMA, etc.) are rising fast — democratizing AI power for smaller startups and solo devs, not just tech giants or governments.
3. The Role of the Creator Economy
Lee talked about AI wiping out jobs — but didn't fully anticipate how AI would empower creators, not just replace them.
Today:
-
Writers use ChatGPT for content creation
-
Solo entrepreneurs automate marketing
-
Students use AI to monetize skills while learning
The book missed how humans and AI would team up, not just compete.
Why This Book Still Matters in 2025
Despite a few misses, AI Superpowers remains an essential read for:
-
Understanding the geopolitical power struggle behind AI
-
Seeing how China’s tech ecosystem works differently from the West
-
Reflecting on the ethics, inequality, and job impact of rapid automation
The book isn’t just about AI — it’s about how societies adapt to exponential change.
Final Verdict
| Category | Verdict |
|---|---|
| AI prediction accuracy | ✅ Mostly correct |
| U.S. vs China analysis | ✅ Still very relevant |
| Generative AI forecast | ❌ Underestimated |
| Workforce disruption | ✅ On point |
| Creator economy vision | ❌ Missed the empowerment side |
Where to Get the Book (Affiliate-Friendly)
📘 AI Superpowers by Kai-Fu Lee – Amazon
🎧 Listen on Audible – Narrated by Mikael Naramore
📱 Kindle Edition
(These are affiliate links — thank you for supporting my blog!)
Your Turn:
Have you read AI Superpowers?
What do you think it got right — or wrong — about the future of AI?
Let’s talk in the comments 👇
Subscribe by Email
Follow Updates Articles from This Blog via Email
No Comments