Every empire looks inevitable — until it isn’t. Rome, the British Empire, Big Oil. Today’s AI titans — OpenAI, Anthropic, Google, Microsoft, and yes, Apple — carry themselves like they’re writing the next chapter of destiny.
But as Karen Hao reminds us in her powerful book Empire of AI, inevitability is an illusion. What looks like technological progress is often empire-building in disguise — extraction of data, energy, and human labour, wrapped in glossy narratives about innovation.
The Empire Lens
Hao’s work forces us to strip back the marketing and see AI as empire:
Extraction: annotation work, content moderation, vast amounts of energy and water.
Concealment: hidden supply chains, silent externalities.
Concentration: decision-making and governance locked in the hands of a few firms.
It’s not just about the brilliance of the models. It’s about the costs — human, environmental, societal — that the empire doesn’t want you to see.
Earn the Right Perspective
In Earn the Right, I argue that technology only works when it serves outcomes that matter — not vanity metrics, not scale for its own sake. Empires collapse because they forget that legitimacy isn’t built on dominance. It’s earned, repeatedly, through competence, accountability, and contribution to human flourishing.
That principle applies just as much to AI. The question isn’t “who can build the biggest model?” It’s: “who can earn the right to shape the future responsibly?”
Judging the Players with Hao’s Risk Matrix
If we take Hao’s empire criteria and apply them to today’s major AI players, the picture gets interesting:
- OpenAI – Innovative, bold, but opaque. Moving faster than its governance can keep up.
- Anthropic – Safety-framed, but still empire logic: massive models, massive compute, centralised power.
- Google – The old empire, adapting. Deep resources but high secrecy and external costs.
- Microsoft – The empire by proxy, using its cloud and capital to extend reach via OpenAI.
- Apple – The outlier. Risk of outsourcing its AI “brain,” but also a chance to chart an anti-empire path: local AI, privacy-first, user-centric. If it chooses courage over convenience, Apple could show another way.
The Human + Technology Interface
This is where my work and Hao’s lens intersect. Empires fail when they forget the human. Strategy execution, with or without AI, is about balancing the interface:
Humans define the purpose.
- Technology amplifies intelligence.
- Outcomes create legitimacy.
- Lose that balance, and you drift into empire. Keep it, and you build systems that endure.
The Future Isn’t Pre-Ordained
The pursuit of superintelligence isn’t destiny. Neither is empire. The future of AI will be written not by GPUs alone but by the leaders and organisations who earn the right to wield this power responsibly.
The critical question is not when we reach superintelligence, but who gets to define what “super” means.
Perhaps now is the time to revisit what we believe our Winning Aspirations should be. Beyond the narrow and predictable targets of quarterly earnings and macro-economic growth. When we align behind different outcomes — human flourishing, resilience, environmental sustainability, fairness — we change the brief. We ask technology to deliver outputs that serve these goals, not the empire’s.
Empires fall. But great systems endure — when they respect the balance between human agency and technological capability. That’s where legitimacy lives. That’s where the future should be.