Is Silicon Valley Becoming an Empire? A Response to Karen Hao’s AI Inflection Point

This morning, I read Karen Hao’s guest essay in The New York Times, titled “Silicon Valley Is at an Inflection Point.”

As someone who frequently reflects on the evolving roles of AI, the tech industry, governments, and civil society, I found the piece intellectually stimulating. Karen argues that the AI sector is no longer just a technological frontier—it is transforming into a modern-day empire. Drawing on themes such as the Trump administration’s AI policies, the narrative around AGI, data extraction, resource dominance, and talent consolidation, she offers a compelling and sobering critique of where Silicon Valley might be headed.

Her essay prompted me to write a few thoughts of my own this morning.

Summary of Karen Hao’s Core Arguments

In her essay, Karen Hao frames Silicon Valley as standing at a historic inflection point—poised to become an AI-driven empire.

She opens with the announcement of the “Stargate Project,” launched by President Trump upon his return to office. The initiative promises $500 billion in private investment over the next four years toward AI infrastructure. For comparison, the Apollo Project—which sent the first humans to the moon—cost the modern equivalent of around $300 billion over 13 years. OpenAI CEO Sam Altman remarked, “It sounds crazy big now. I bet it won’t sound that big in a few years.”

Karen sees this governmental pivot as more than symbolic. The Trump administration has made clear that it intends to back the tech industry without reservation. A Republican tax bill recently passed in the House includes provisions to block state-level AI regulation for the next decade. This full-throttle federal endorsement, she argues, significantly expands tech companies’ influence and insulation from oversight.

According to Karen, AI companies in Silicon Valley are no longer just technology firms. They are entities that acquire land, create their own currencies, reshape economic structures, and even influence political decisions—acting increasingly like sovereign powers. In her view, these are the hallmarks of empire.

The central ideological vehicle of this "empire" is the promise of AGI—Artificial General Intelligence. Under the banner of advancing humanity, tech CEOs portray their missions as morally essential. Anthropic’s CEO, Dario Amodei, declared that “AI from the U.S. and democratic nations must be better than China’s—otherwise we lose.” AGI, then, becomes not just a technological goal but an ideological shield.

This rhetoric, Karen warns, justifies sweeping data extraction and unchecked resource consumption. For instance, Meta—despite already possessing data from nearly 4 billion accounts—has reportedly scraped websites indiscriminately and even considered acquiring publishing giant Simon & Schuster to train its AI. These practices raise serious concerns about privacy and intellectual property rights.

On the environmental front, AI infrastructure comes at a staggering energy cost. According to early Stargate Project drafts, its supercomputers could consume the same electricity as 3 million homes. McKinsey projects that by 2030, global power grids would need to add capacity equivalent to 2–6 times California’s 2022 energy use if AI expansion continues at its current rate. One OpenAI employee reportedly told her: “Now we’re short on electricity and land.”

Karen also highlights the consolidation of AI talent. In 2004, only 21% of AI PhD graduates entered private industry. By 2020, that number had soared to around 70%, many lured by compensation packages exceeding $1 million. As a result, critical voices have been absorbed into the very companies they might once have scrutinized.

One example she cites is the dismissal of Google’s Ethical AI team leaders shortly after publishing a paper warning about the dangers of large language models (LLMs)—the foundation of tools like ChatGPT. In Karen’s view, the industry is becoming intolerant of internal critique.

Ultimately, she warns that this AI “empire” could centralize power to such an extent that data, energy, talent, and political discourse are monopolized—leaving citizens and democratic institutions increasingly sidelined.

Is Silicon Valley Really an Empire? A Few Counterpoints from LexSoy

Karen Hao’s essay offers rich insights and timely warnings. However, I believe her narrative risks being overly one-sided and deterministic. By casting Silicon Valley as an inevitable empire, she overlooks the complex, evolving checks and balances within both the tech industry and its regulatory landscape.

1. Is “Empire” a Fair Analogy?

It’s true that tech companies hold enormous global influence. But to describe them as empires—akin to sovereign states with unilateral authority—feels like an overextension.

For example, OpenAI has maintained a nonprofit parent structure and a capped-profit model, preserving some insulation from total investor control—even as Microsoft holds a significant stake. Moreover, legal guardrails exist. The EU’s AI Act, FTC antitrust probes, and state-level privacy laws such as California’s CCPA all serve to limit corporate overreach.

Silicon Valley remains subject to competitive markets, civil pressure, and democratic lawmaking. It is powerful, yes—but not unaccountable.

2. AGI: Religious Dogma or Strategic Vision?

Karen characterizes the AGI narrative as quasi-religious. But I would argue it’s better understood as a strategic vision—designed to attract capital, talent, and public interest.

Technological grandiosity is not new. We saw it with cloud computing, smartphones, and self-driving cars. While AGI may not arrive as promised, it serves as a rallying concept—spurring discourse and investment.

At the same time, non-AGI models like Google DeepMind’s AlphaFold (which helps predict protein structures) show that AI progress isn’t exclusively focused on general intelligence. The field is broad and pragmatic, and internal recognition of AGI’s limits is growing.

3. Does Industry Talent Absorption Eliminate Accountability?

Karen links the influx of researchers into industry with a loss of independent oversight. But this view underestimates the internal checks now emerging within tech companies.

Meta, Anthropic, and DeepMind all maintain dedicated AI safety teams. Anthropic, for instance, has invested heavily in constitutional AI—embedding ethical principles directly into models. OpenAI has explored restructuring its board to include ethics specialists.

Internal governance, while imperfect, can sometimes outperform external watchdogs—especially given how slowly public regulation tends to catch up.

4. Resource Consumption Without Context

AI’s energy demands are real. But so are those of many other industries: manufacturing, shipping, streaming, crypto mining.

Crucially, AI is also part of the solution. Google AI, for instance, has developed systems that reduce energy consumption in data center cooling, saving hundreds of thousands of kilowatt-hours annually.

If we criticize AI solely for energy use, we must also scrutinize the smartphones, YouTube videos, and remote work tools we rely on daily. The question is not whether AI consumes resources—but whether it delivers sufficient societal value in return.

5. Legal and Regulatory Dynamics Are Still in Play

Finally, the proposed federal ban on state-level AI regulation that Karen mentions is far from law. Given America’s federalist structure, such a move would likely face substantial legal and political hurdles.

States like California, Illinois, and Colorado continue to legislate consumer protection around AI. The FTC, meanwhile, has issued a steady stream of guidance on AI transparency and algorithmic fairness since 2024.

Far from a regulatory vacuum, the current environment is one of overlapping legal frameworks and active negotiation—not empire-building through impunity.

In Closing: Technology Doesn't Become an Empire on Its Own

Karen Hao’s essay raises urgent and necessary questions. The rapid growth of AI, especially when coupled with state support and weak oversight, poses real risks—to privacy, labor, the environment, and democratic accountability.

But portraying Silicon Valley’s future as a linear march toward empire oversimplifies reality. In truth, the direction of AI governance remains undecided—and contested.

Tech companies are still operating within regulatory and market constraints. AGI is still debated, not decreed. Internal ethics structures are evolving. Civil society, policymakers, and international institutions are increasingly engaged.

We should focus less on fear-based framing and more on building the right frameworks. Transparency, accountability, education, ethical design, and enforceable policy are the tools we need.

Silicon Valley may be at an inflection point—but that turn will not be taken by tech firms alone. It will be shaped by governments, researchers, advocates, and yes, users.

Technology doesn’t become an empire by itself.
It only does when we stand back and watch it happen.

© LexSoy Legal LLC. All rights reserved.
All content on this website is the intellectual property of LexSoy Legal LLC and is protected under applicable copyright and intellectual property laws. Reproduction or redistribution without express permission is prohibited.