Part I — The Mirage of Innovation

By Lorde Astor West 10/12/2025

For much of the past decade, the tech industry has relied on the illusion of acceleration. Artificial intelligence, once framed as a moonshot, has become the most powerful story in the market—an idea so large it can justify any valuation.

(from “Who Owns the Benefits — and Who Pays the Cost?”)

At the dawn of 2025, Sam Altman declared that OpenAI had “figured out AGI.” The statement rippled through the global press like a starting gun. Stocks surged. Analysts recalibrated projections. Commentators debated the ethics of an intelligence that, supposedly, could now understand anything a human could.

But as the year winds down, the promised “general intelligence” has materialized into something far more ordinary: a suite of commerce integrations, video-generation tools, and an in-app shopping feature inside ChatGPT. The world didn’t get a new species of mind; it got another app store.

It’s a familiar bait-and-switch in Silicon Valley’s long history of innovation theater. When genuine breakthroughs stall, monetization steps in to fill the silence.

The Economy of Illusion

For much of the past decade, the tech industry has relied on the illusion of acceleration. Artificial intelligence, once framed as a moonshot, has become the most powerful story in the market—an idea so large it can justify any valuation. The narrative is self-sustaining: announce a leap, raise a round, spend on infrastructure, and repeat.

But under the surface, growth has decoupled from progress. Many of the companies driving the “AI revolution” are not profitable in any traditional sense. Their revenue is recycled capital—venture dollars funding startups that, in turn, buy compute time from the same cloud providers that financed them. It’s a closed circuit of liquidity masquerading as market demand.

The hyperscalers (AWS, Azure, Google Cloud) are also their own best customers.They rent GPUs to themselves, fund AI startups through venture arms, then report that revenue as cloud growth.This creates circular demand — money recirculating inside the same ecosystem, like counting your reflection as a new customer.

In the 1990s, this was called the dot-com feedback loop. In 2025, it’s simply called “AI adoption.”

When the Frontier Becomes a Franchise

The rhetoric of intelligence and transformation persists, but the product pipeline tells a different story. The same laboratories that once promised to “solve intelligence” are now launching features indistinguishable from those of mainstream consumer tech: embedded commerce, personalization widgets, branded assistants.

To outside observers, this looks like momentum. To insiders, it feels like stagnation. The cutting edge of machine learning has plateaued; transformer models—powerful, but increasingly inefficient—are reaching diminishing returns. Yet instead of pursuing new architectures, companies are optimizing for distribution. They’re not building smarter systems; they’re building stickier ecosystems.

Innovation has become a defensive maneuver—less about discovery, more about maintaining control.

From Intelligence to Infrastructure

Even the language has shifted. In research circles, “advancement” used to mean conceptual breakthroughs. Now it refers to scaling efficiency, latency improvements, or power utilization. In corporate decks, “AI innovation” often translates to “API integration.”

This semantic drift signals something deeper: a transition from scientific pursuit to infrastructure monopoly. The goal is no longer to understand intelligence—it’s to own the pipelines through which everyone else accesses it.

It’s less about creation, more about colonization.

The Real Invention Was the Business Model

What the world is witnessing isn’t the arrival of artificial general intelligence—it’s the financialization of artificial intelligence. The brilliance is no longer technical; it’s economic. By packaging compute as intelligence, and intelligence as subscription, the industry has found a way to monetize speculation itself.

Each new model release sustains the illusion of progress long enough to feed another round of investment, another GPU order, another infrastructure buildout. The system keeps moving, even if the technology isn’t.

There’s no collapse yet—just a shimmering heat haze of optimism over an increasingly barren landscape of real innovation. It looks like motion. It feels like progress. But it’s mostly inertia.

In hindsight, this may be the defining paradox of the AI era: a moment when humanity poured unprecedented resources into teaching machines to think— and ended up training itself to believe.