The Rigged AI Game You Can’t Unsee (Once You See It)

The real cost of AI isn’t "time saved" but "power lost".

Credits: Diana

The "paper clip problem" is a classic thought experiment in AI ethics. It imagines a superintelligent AI given one goal: to make as many paper clips as possible.

Without common-sense limits, the AI pursues this goal so relentlessly that it transforms all of Earth’s resources—and humanity itself—into components for a planetary paper clip factory.

When I first started using generative AI, I was an evangelist. Like many, I devoured tutorials, my brain buzzing with a dopamine hit of new possibilities. It felt like a cognitive superpower.

But my focus shifted after reading Karen Hao’s new book “Empire of AI” and the AI Now Institute’s recent report“Artificial Power.” I started to ask myself whether I am using AI for genuine progress, or feeding technology that makes exploitation more efficient.

Are we already living in an unfolding reality of the paper clip problem?

The small, individual glitches, like the weirdly agreeable answers, look less like bugs and more like features of a much larger, more troubling system. My question is changing from “How can I use this?” to “What is this system I'm using?”

That’s why this exploration is not about ‘AI Prompts for learning,’ but a more zoomed-out take, viewed through a lens of power and justice. My goal isn’t to give you answers, but to help us ask better questions.

---

You’re Playing in an Intentionally Rigged Casino

The current AI boom is not neutral technological progress. It is a strategic and aggressive concentration of power. This sentence from the AI Now Institute that kept me pondering:

"The question we should be asking is not if ChatGPT is useful or not, but if OpenAI's unaccountable power, linked to Microsoft's monopoly and the business model of the tech economy, is good for society."

How I read this is twofold.

First, the current form of AI (Large Language Model Chatbots like GPT & agents) was not inevitable. It is the result of thousands of choices made by a handful of homogeneous actors (Silicon Valley elite).

Second, those choices are not accounted for, and we have no mechanisms in place to restrict, audit, or democratically govern the concentrated power of the firms, making them, despite their sweeping impact on our institutions, labor markets, public resources, and fundamental rights.

Karen Hao writes that these power holders act like colonial empires. They seize and extract resources that are not theirs—our collective data, our art, our writing—and exploit unseen labor to refine those resources for their enrichment.

What comes to my mind is a rigged casino, an economic system designed so the house always wins. Here’s the AI game I can no longer unsee:

1. The House Owns the Infrastructure.
   Big Tech (Microsoft, Google, Amazon) owns the cloud platforms where nearly every AI tool runs. Every startup, every prompt, pays them rent. This dominance over compute and infrastructure gives them gatekeeping power—not just over who builds AI, but over what kinds of AI get built, and whose values shape it.

2. The House Sells the Hype.
   The leading AI players manufacture urgency with what the AI Now Institute calls "AI's False Gods"—grandiose narratives about an "AI arms race" or imminent Artificial General Intelligence that serve as "the argument to end all arguments," justifying deregulation and sidelining debate. By claiming that if they don’t build AGI, someone else (often framed as a geopolitical adversary like China) will, it frames any critique or delay as dangerous. It keeps public discourse fixated on speculative futures rather than the very real, present-day harms of AI systems

3. The House Hides the Losses.
   The financial ground beneath this hype is incredibly shaky. The AI Now report highlights that the industry is burning through cash at an astonishing rate. OpenAI, for example, has lost half a billion dollars in one year. These sky-high valuations hinge on an expensive technology that still lacks a proven, profitable business model.¹ And yet, big AI players sideline other possible directions for AI—such as smaller, community-driven, or publicly governed models—by monopolizing attention, funding, and infrastructure.

4. The Public Foots the Bill.
   But “the public” isn’t a monolith. The costs fall unevenly—on low-income communities living near polluting data centers, on workers subjected to algorithmic surveillance, and on people across the Global South whose data, labor, and environments are extracted to fuel systems they didn’t ask for and rarely benefit from. Meanwhile, a concentrated elite reaps the rewards.

---

The Misdirection

The promise is that AI will make you more efficient and create widespread prosperity.

But the evidence points elsewhere. Karen Hao referenced analysts Parmy Olson and Carolyn Silverman, who concluded, this technology may never deliver on its economic promise, but instead "just concentrate more wealth at the top.”

The productivity and prosperity narrative is the ultimate misdirection. It keeps us focused on our own small, individual gains, blinding us to the enormous societal costs and the systemic power being amassed.

It works because, as Hao notes, the people building this technology are masters at framing whatever they have to offer in terms of exactly what you want.

---

Resist the Narrative

Artificial intelligence doesn’t have to be what it is today. We can, as Hao urges, resist the narratives that OpenAI and the AI industry have told us.

The current paradigm—rooted in domination, extraction, and profit—is not inevitable. It is only one path. And by naming it, we make space for other alternative paths that might serve our society’s real needs.

I can’t offer a quick fix. But I hope that now you’ve seen the contours of this rigged system, you can’t unsee them.

It's time to demand better.

Because silence will not protect us.

1. A critical dimension less explored in these otherwise excellent analyses is the rapid integration of AI into military intelligence and state surveillance. The use of AI-powered targeting systems from Ukraine to Palestine represents a significant escalation. This omission is notable because the military-industrial complex is a primary, and highly profitable, frontier for AI development. The competition between firms is therefore not merely about market share, but about who will become the dominant supplier of state power.