This video takes the 2025 book "Empire of AI" as its foundation to accessibly unpack the hidden essence of "power" and "inequality" in the AI industry. It examines how OpenAI and Big Tech companies are building a new form of imperialism through AI, and highlights the issues we need to grapple with.


1. Video Introduction -- About the Book and the Author

The video opens casually by recommending "Empire of AI" as a great vacation read. Despite its weighty title, the book was published in May 2025 and became a New York Times bestseller as a serious piece of tech journalism.

The reviewer introduces the book as an in-depth investigative account centered on OpenAI and Sam Altman, covering the AI development processes of Big Tech companies. He emphasizes that the author conducted over 300 interviews with 260 people, 150 of which were with OpenAI employees.

"This is what you'd call a quintessential tech journalism book."

Author Karen Hao studied mechanical engineering at MIT, served as a top editor covering AI and technology at MIT Technology Review, and has contributed to the Wall Street Journal and other outlets. She has been deeply following and reporting on OpenAI since 2019.


2. Key Content -- Big Tech's Story, OpenAI's Rise and Transformation

Before diving into the full summary, the reviewer notes that "Empire of AI" isn't just about OpenAI -- it also broadly illuminates the power structures and capital logic of the AI era.

"While it tells the internal story of OpenAI, it says much more than that. It documents how scientific ambition became ideological and capital-driven."

The book first traces the history of AI industry development. Initially, pure algorithmic research was vibrant, but over time the game shifted to a "brute force" approach -- massive models, enormous data, and staggering computing resources became the deciding factors.

"At some point, the brute force game began. Building huge models with massive parameters, trained on enormous amounts of data using incredible computing resources. That's what the game became."

"OpenAI is the company that took this game to the next level."

Around 2017, OpenAI internally concluded that "the key to reaching AGI (Artificial General Intelligence) is computing power" and chose a resource-intensive strategy. This was partly because Google was technically ahead, and OpenAI needed a different approach to catch up. Ultimately, this strategy propelled OpenAI to the top tier of Big Tech, but it also brought the side effect of the AI industry requiring massive electricity consumption and carbon emissions.


3. Three Major Impressions from the Book

The reviewer shares three powerful impressions:

  1. The book's immersive quality and factual rigor Despite reading like a novel, every detail is backed by field reporting and interviews.

    "It was genuinely fascinating. Even though it's investigative journalism and not fiction, I couldn't stop reading."

  2. A new and deeper understanding of the AI industry Hidden dynamics and untold stories within the AI industry are revealed in detail.

  3. The imperialist nature of the AI industry The most sobering takeaway was the book's argument that the AI industry resembles historical imperialism. The reviewer outlines the author's case in four main points.


4. Four "Imperialist" Characteristics of the AI Industry

As the book's core message, it critiques the AI industry as resembling past imperial exploitation.

  1. Unauthorized appropriation of others' resources AI uses other people's data -- others' creative works and information -- as training material without permission.

    "The AI industry trains AI models using data it doesn't own."

    "These trained AIs end up taking away the jobs of the very people who created the data."

  2. Exploitation of economically vulnerable populations Basic work like data labeling and content filtering relies on cheap labor in low-income countries in Africa and Latin America, without proper compensation or support.

    "AI companies are using people from economically vulnerable societies in Africa and Latin America for data labeling, but compensation and support aren't provided systematically."

    "The structure for maximizing profits is no different from past imperialist policies."

  3. Monopolization of knowledge production As AI researchers get absorbed into large corporations, pure research declines, and more research focuses solely on commercial product development and revenue.

    "Because AI companies are absorbing so many AI researchers, the focus shifts from pure AI research to product-oriented research."

  4. The "good empire vs. evil empire" logic The tendency for each side to claim only their own camp is legitimate -- a "rules for me but not for thee" approach to dominance -- closely mirrors imperialism.

    "They approach it with a good empire vs. evil empire logic: we're the good ones, they're the bad ones, so we should hold the reins."

    "The U.S. says technology must be protected. OpenAI says it's dangerous for Google to have AGI. They decide for themselves who's good and who's evil."


5. The Reviewer's Thoughts on the Author's Arguments

The reviewer largely agrees with the author's arguments but honestly shares that he felt some uneasiness and reflection about directly comparing the AI industry to the history of imperial colonialism.

"Honestly, directly comparing imperialist history to the AI industry was a bit uncomfortable."

"The history of imperial colonial rule is incomparably heavier and more tragic than the injustices of the AI industry. I worried it might inadvertently trivialize that history."

However, he evaluates the book as deeply meaningful for encouraging us to think about AI not merely as a technological matter, but in terms of its implications for human rights, history, and inequality.

"The significance was in making me look beyond the technology to see the human and resource dimensions, and even reflect on history."


6. Closing -- A Strong Recommendation

The video wraps up with the reviewer sharing that the book was so absorbing it made him forget the summer heat, and he strongly recommends it to viewers.

"It was a truly immersive and enjoyable book. I definitely recommend everyone read it!"


Conclusion

"Empire of AI" incisively examines the hidden power structures, inequality, and imperialist attributes of AI. It is not just a technology explainer but a book that makes us deeply reconsider where we're headed with AI -- a thoughtful guide to the critical issues of our time.

Related writing