Like many of you, I am incredibly passionate about the potential of Artificial Intelligence. I spend my days exploring new tools, experimenting with prompts, and imagining how this technology can transform the way we work and create. It often feels like pure magic.
But recently, I realized I needed to look past the magic and understand how the trick is actually performed. I picked up Karen Hao’s investigative book, Empire of AI: Dreams and Nightmares in Sam Altman's OpenAI.
I’m not a machine learning engineer or a world-class AI researcher; I’m just an enthusiastic explorer of this space. But reading this book was a profound, grounding experience. Hao’s meticulous research pulls back the curtain on the race toward Artificial General Intelligence (AGI) and reveals the stark physical and human realities powering our favorite tools.
If you are fascinated by AI, this is a must-read. It challenged my optimism with a necessary dose of reality—and ultimately made me realize that the future of this technology shouldn't just belong to the tech giants. It belongs in our classrooms, our workplaces, and our learning communities.
Here are the hard truths the book forced me to confront, and why I believe the learning space and education is our way forward.
We Are Passengers in an Arms Race
When we log into an AI platform, we experience the results of brilliant innovation. But Hao’s reporting highlights that the current obsession with building massive Large Language Models isn't just a natural scientific evolution—it’s an arms race driven by fear and game theory.
The mantra among the top players has become "scale or die." The race to AGI is increasingly about dominating the market and securing billions in funding, rather than thoughtfully solving human problems. It made me realize that the current path we are on isn't the only way AI could have developed; it’s simply the path chosen by a few powerful voices who tell the most compelling stories to investors.
The "Cloud" is Heavy, Hot, and Thirsty
It is so easy for us to forget that the digital world has a physical footprint. Hao brings us back to earth, detailing the staggering environmental toll of training and running these "monstrosities" (as she calls generative AI models).
Compute demand is growing exponentially, while our energy and water supplies are finite. The book documents heartbreaking stories—like data centers threatening the drinking water of drought-stricken communities in Chile. The AI we use to draft emails and brainstorm ideas requires massive, water-guzzling, power-hungry infrastructure. Acknowledging this physical ceiling isn't being anti-tech; it’s a necessary reality check for anyone advocating for a sustainable future.
The Invisible Humans Behind the Machine
This was the most sobering part of the book. When our AI assistants reply to us politely and safely, it’s not because the algorithm is inherently moral. It’s because an invisible workforce—often in the Global South—has been paid poverty wages to filter through the darkest, most traumatic content on the internet to train it.
Hao draws a powerful, uncomfortable parallel to historical empires: extracting raw resources (in this case, data and cheap labor) from vulnerable populations to consolidate wealth and power at the top. As someone who advocates for AI's potential to uplift people, confronting this "AI colonialism" is a bitter pill. We cannot celebrate a tool that benefits us if its foundation relies on the exploitation of others.
Why "Free Money" Won't Fix the Meaning Crisis
Silicon Valley often pitches Universal Basic Income (UBI) as the ultimate safety net for the AI revolution: If AGI takes our jobs, the tech companies will just pay us to live.
But Empire of AI points out the deep flaws in this thinking. Humans are wired for purpose. We need the identity, social connection, and rhythm that learning and contributing provide. If a massive swath of the workforce suddenly finds their cognitive skills replaced by a prompt, the resulting crisis of meaning cannot be solved by a monthly stipend. We don't just want to survive; we want to matter.
The Call to Action: Why Learning and Development Is the Key Player to Take the Wheel
It’s easy to read these realities and feel helpless. If the "Empire of AI" is backed by billions of dollars and massive server farms, what can normal, everyday explorers like us actually do?
This is where my reflection shifted from concern to a clear sense of mission. The antidote to an empire of concentrated power is the democratization of knowledge. Our greatest lever is education.
We cannot leave the future of AI solely in the hands of a few tech billionaires. We need Learning and Development (L&D) professionals, educators, and community leaders to step up. If we want to collectively influence how AI integrates into society, we must focus on:
- AI Literacy over Blind Adoption: We need to teach people not just how to use AI, but how it works, where its biases lie, and what its costs are. Critical thinking is one the ultimate future-proof skills.
- Human-Centric Upskilling: Instead of bracing for replacement, L&D should focus on cultivating the skills AI cannot replicate: deep empathy, complex problem-solving, ethical reasoning, and community building.
- Empowering the Workforce: By educating everyday people on the realities of AI, we empower them to demand better policies, advocate for sustainable tech, and build smaller, community-driven models that solve local problems without exploiting global resources.
Empire of AI taught me that the future is not predetermined. The technology is here, and it is incredibly powerful. But how it shapes our world is a human choice. By committing to continuous learning, sharing our knowledge, and fostering a culture of critical, compassionate education, we can ensure the AI revolution becomes a tool that elevates all of us, rather than an empire that rules us.
