Surviving the AI Era | Vol. 22
[Vol. 22] Co-Intelligence Is Already Outdated
I gave my best at work for over 20 years. My life is still tough.
Living under capitalism, I never once thought about owning capital. I just sold my labor and time, believing that would be enough to get by. That delusion is exactly what made my life what it is today. I only realized this now.
Now everyone's talking about AI. I don't know much about it, but the fear of repeating the same mistake I made for 20 years—working hard while staying ignorant—has me watching every AI video I can find.
Sam Altman, Elon Musk, big-name executives talking about the future of AI. Professor Kim Dae-sik's various takes. It was fascinating. It was new. But at some point, the excitement faded.
Before I knew it, my question had changed from "Wow, that's amazing" to "So what am I supposed to do?"
YouTube doesn't really have answers to that question.
So I turned to books. Short YouTube clips can spark a moment of thought, but to actually build my own perspective, I needed something deeper—a book that someone spent years struggling to write. That's when I started reading AI books.
💡 Same AI. Completely different stories depending on who's talking.
· Ray Kurzweil (Scientist)
"The Singularity—the moment AI surpasses humanity—is coming."
· Ethan Mollick (Business Professor)
"Use AI as your second brain to expand your capabilities."
· Yuval Noah Harari (Historian)
"AI is dangerous."
· Yudkowsky & Soares (AI Safety Researchers)
"Humanity goes extinct."
I wanted a roadmap for the future of AI, so I first read Ray Kurzweil's The Singularity Is Nearer.
The second book I picked was Ethan Mollick's Co-Intelligence (Korean title: Dual Brain).
I chose it to learn the practical "how-to" of using AI.
After this, I plan to read Yuval Noah Harari's Nexus for a historian's take on AI, then AI: The Birth of God, The End of Humanity by Yudkowsky & Soares—the worst-case scenario.
But first, this is a review of Co-Intelligence, so—
Bottom line: In 2026, there's no need to read this book.
I Trusted the Recommendations
YouTube recommended it. AI recommended it when I asked.
New York Times bestseller. The Economist's Book of the Year for 2024. Korean edition published March 2025.
But the whole time I was reading, one question wouldn't go away: 'Why is everyone still recommending this?'
This book was written in 2023. When GPT-4 had just come out.
It's now May 2026. In the AI world, three years isn't one generation. It's three.
Countless YouTubers are still promoting this book. Whether they actually read it or got paid to do it, I have no idea.
But one thing is clear. If you're already using AI, you won't find anything new here.
Why It Didn't Click
I hesitated before buying it. The original came out in 2024, Korean edition in March 2025. Felt a bit old for 2026.
Mollick debates with ChatGPT in the book. "Do you actually think? Do you have feelings?"
He asks AI to tell jokes. He teaches how to write good prompts.
He suggests ways to use AI at work.
In 2023, this was probably cutting-edge stuff.
But right now, I'm running three AIs simultaneously. Claude, ChatGPT, Gemini.
While Mollick was talking about a "Dual Brain," people are already running triple.
I kept hearing that "prompt engineering" is the key skill for using AI well.
Honestly, I have no idea why prompt engineering is supposed to be such a big deal.
It's just conversation. Say what you want clearly, and you get a clear answer.
At work, the most annoying thing a boss can say is "just handle it somehow."
Give AI a vague prompt, you get a vague answer. That's all there is to it.
In 2023, they called this "engineering."
In 2026, shouldn't we just call it "communication" or "being clear"?
If you pay enough for a premium subscription, the AI remembers your ongoing work. In the same chat thread, what used to take paragraphs of explanation now takes a single sentence.
Me
"Let's start the daily routine. Go ahead."
AI
"Got it. Same workflow as always."
⏱️ A bit later, it gets even simpler:
AI
"On it." And the work starts immediately. Dead accurate, every time.
Academic Optimism, Real-World Cynicism
There's one part of this book that really got to me.
Mollick says:
"When AI takes over repetitive tasks, humans can focus on more creative work. Productivity goes up, and we get time for more meaningful things."
But what actually happened in the three years since?
AI does the work well enough that three people are no longer needed. One is enough. Two are gone.
Creative work? Meaningful endeavors?
No. Two people lose their livelihoods, and the one left standing does the work of three.
Mollick turned out to be just that—a business school professor, not a CEO running an actual company.
The Second Brain Isn't Free
There's another thing Mollick left out.
Dual Brain—sounds great. An era where everyone gets a second, AI-powered brain.
Except that second brain requires a monthly subscription.
Claude, ChatGPT, Gemini. Subscribe to all three and it's around $50–60/month.
Add AI agents and coding tools, and it's hundreds.
Forty years ago, people lived fine without phones.
Today, you can't function without one.
Phone bills became a new baseline cost of living.
AI is heading down the same path.
But phones got cheaper as they became mainstream.
With AI, the more useful the feature, the more expensive it gets.
People with money run premium AI and write code with it.
People without money use the free tier to ask simple questions.
This time, money directly determines the gap in individual capability.
The definition of "basic living costs" is getting blurry.
It used to be food, shelter, clothing.
Now it includes phone bills, internet, and soon, AI subscriptions.
Nobody can draw the line anymore.
Mollick said "upgrade to a Dual Brain with AI."
But people who can't afford it stay Single Brain.
Mollick doesn't talk about this part.
So What Is Mollick Saying Now?
Throughout the book, Mollick stays relentlessly positive about human-AI collaboration.
It made me a bit angry, honestly.
Three years later, people are losing jobs to AI and anxiety is everywhere.
Does he still believe that rosy picture?
Or has he started to question his own assumptions? I got curious.
So I asked AI to dig up his latest papers, articles, and talks.
Sure enough, he's been changing too.
· 2023: "AI is humanity's partner." (Co-Intelligence)
· Late 2025: "Honestly, things are very different now from back then."
· 2026: His new book is titled Co-Existence.
From Co-Intelligence to Co-Existence. From "let's work together" to "let's survive together."
The title says it all.
And he's started saying things like this:
"The goal of AI companies is to build machines smarter than humans and replace human labor. It might be unrealistic, but it can't be ruled out."
In the last section of his book, he cautiously raised the possibility—an era where AI builds AI, superintelligence.
Back then he wrote about it like a distant hypothetical. Now he seems to be admitting it's real.
Honestly, I don't know. Should we even call business management a real academic discipline?
If you're a scholar, shouldn't you be more careful when you're talking about things that affect people's lives?
Mollick's book must have influenced a lot of people—executives, AI developers, the general public. Did it, without anyone noticing, help fuel the unstoppable AI arms race we're stuck in now?
Was It a Waste of Time?
It's not a bad book. If I'd read it in 2023, it would've blown my mind.
But if you're already using AI instead of Google to get information and handle daily tasks, this book is a waste of time.
On the other hand, if you've never once asked AI a question, it might be worth a read.
The Journey Isn't Over
In the end, this book was just... a pile of text I wanted to get through as fast as possible.
The third book is waiting: Harari's Nexus. How does a historian see the AI era?
The fourth: Yudkowsky's AI: The Birth of God, The End of Humanity. An extreme warning from AI safety researchers that I never saw coming.
And once I finish all four, I think a fifth question will come.
Companies will keep racing to dominate AI, no matter the cost. They won't care much about warnings or safety concerns. When that happens, the only thing we can count on is government policy.
How do we fix wealth inequality in an automated world? AI-era government policy. When you need to eat to survive, that's where your attention inevitably goes.
For now, I want to see how Harari—the historian—looks at this AI era. Nexus is next.
