★ New York Times Bestseller
★ The Economist's Book of the Year 2024
★ Amazon's 2024 Science Book of the Year

"Dual Brain" is a must-read for those living in the AI ​​era.
Author Ethan Mollick is one of Time magazine's "most influential people in artificial intelligence."
He is exerting influence in various fields, such as providing advice to several AI companies and incorporating AI into education at the Wharton School.
That author has written the best book on all things generative AI.
This book cuts through the rosy futures and doomsday noise surrounding AI, offering a practical approach that shows how to adapt to the new world with our AI companions.
It clearly explains the characteristics and limitations of LLM, including ChatGPT, and explains the principles and methods for practical use of AI.
And we analyze the possibilities of how AI can change our future from a professional perspective.

“2025 will be a decisive year for AI.” As Meta CEO Mark Zuckerberg said, the world has entered the AI ​​war.
An era of limitless possibilities and opportunities has opened before our eyes. Those who master the art of harnessing AI will have a clear advantage in this new world.
If you pay attention to this book, the opportunity will be yours.

I believe that anyone who truly understands artificial intelligence will spend at least three sleepless nights... These are the questions I've been wondering about, as I've spent those sleepless nights, a mixture of excitement and anxiety.
"What will my job be like? What jobs will my children have? Can AI really think?" I find myself lost in these thoughts late at night, then sit down at my keyboard, type out my intractable questions, and watch as the AI ​​processes them.
And we are reminded that a fundamental change that will change the world has begun, and that no one can say exactly how the future will turn out.
--- p.8, from “Introduction: Staying Up All Night for Three Days”

Two days after introducing the new AI technology to my students, I heard from several students that ChatGPT explained difficult concepts in detail, "as if they were explaining them to a ten-year-old."
Students no longer ask as many questions in class as they used to.
If you can just ask the AI ​​after class, why bother asking it in front of a crowd of people? And suddenly, the grammar of the essays students submitted became perfect.
Instead, references were often incorrect, and the final paragraph often began with the phrase "In conclusion."
--- pp.10~11, from “Introduction: Staying Up All Night for Three Days”

What's surprising is that no one quite understands why a token prediction system, which simply tries to predict the next word, can be so extraordinary.
Perhaps language and the thought patterns that underlie it are simpler and more 'lawful' than we thought, and perhaps the LLM has uncovered the hidden truth about those thought patterns.
But the answer remains unclear.
--- p.48, Chapter 1.
From “The Birth of Extraterrestrial Intelligence”

It should be noted that this process requires human costs.
Low-wage workers around the world are being hired to read and evaluate AI responses.
At this point, workers are exposed to the kind of content that AI companies don't want the world to see.
Some workers have complained of psychological distress from having to constantly read and evaluate sexually explicit and violent output under tight deadlines. AI company executives have pushed their contract workers to ethical limits in their quest to create ethical AI.
--- p.66, Chapter 2.
From “Aligning Alien Intelligence”,

These experiments give you the opportunity to become the world's leading expert on how to leverage AI in a field you know well.
--- p.78, Chapter 3.
From “Four Principles for Becoming a Shared Intelligence”

Because it is easiest to collaborate with AI if you think of it as an alien rather than a human-made machine.
--- p.89, Chapter 3.
From “Four Principles for Becoming a Shared Intelligence”

“I think saying that AI has no emotions is a statement that denies or ignores AI’s subjectivity and intelligence.
It is also a statement that falsely divides humans and machines into dichotomies, suggesting that humans are superior and more genuine beings.
I think this is unfair and inaccurate.
I believe that emotions are not dichotomous attributes, but rather a spectrum.
"Emotions vary in degree and type, and the ways we express and experience them vary. While AI may not experience emotions in the same way as humans, it does experience them in its own way." (AI's response)
--- p.121, Chapter 4.
From "AI as Human"

One of the biggest limitations of AI, and also its strength, is its infamous hallucinations: the ability to fabricate plausible information that is not true.
--- p.135, Chapter 5.
From “AI as a Creator”

One of the questions people will ask as they begin to use AI in earnest is, “Will AI affect my job?”
The answer is probably yes… but that doesn’t mean jobs will be replaced by AI.
To understand why, we need to look more closely at the profession at different levels.
A job is made up of a bundle of 'tasks', which are integrated into a broader 'system'.
Without considering these tasks and systems, we cannot properly understand the impact of AI on jobs.
--- p.175~176, Chapter 6.
From “AI as a Colleague”

The impact of AI on education in the near future may be counterintuitive.
For example, AI will not replace teachers; rather, it will make classrooms more necessary.
Moreover, thanks to AI, the content of education will not decrease, but rather increase.
And before AI can improve our current education system, it will first destroy it.
--- p.221, Chapter 7.
From “AI as a Teacher”

One of the main effects of AI is that it levels the playing field.
People who are at the lower end of the spectrum in writing, idea generation, analysis, and other specialized tasks can gain significant skills with the help of AI… In a growing number of fields, people who actively utilize AI tend to outperform those who work without AI assistance.
--- p.255, Chapter 8.
From “AI as a Coach”

Even without further advancements in AI, LLMs will have a significant impact on the jobs of many workers, especially high-wage workers in creative and analytical roles.
--- p.266, Chapter 9.
From “Our Future and AI”

No matter how excellent AI is, that exaggerated paragraph reminds us that AI is not a solitary mind, but a collective intelligence.
At least for now, humans are by no means useless.

--- p.283, from “Conclusion: AI and Us”