Technological revolutions are not new. The steam engine replaced muscle power, and the computer automated routine calculations. Each time, society adapted. Jobs changed, but the fundamental economic loop held: people earned wages for their labor and used those wages to buy goods and services.

Artificial intelligence presents a different kind of shift. The 'Discontinuity Thesis' argues that AI is not just another tool. It represents a fundamental break from the past because it automates cognition itself. This isn't about replacing physical tasks, but the very process of thinking, planning, and creating that has been the exclusive domain of human labor. This change challenges the traditional wage-demand circuit that powers modern economies.

The Discontinuity Thesis is built on three core ideas. Together, they paint a picture of a rapid, irreversible economic transformation.

Let's examine each of these premises more closely.

The first premise is about efficiency. When an AI is paired with a human operator, the cost to produce a unit of cognitive work—like a market analysis report, a piece of software, or a legal brief—plummets. This isn't a small improvement; it's a change of an order of magnitude or more.

A human expert requires a salary that covers their living expenses, or subsistence costs. An AI's costs are primarily for computation and energy, which are vastly lower and continually decreasing. The result is a system that can outperform a human-only workflow on both price and speed.

As the table shows, the AI-human hybrid doesn't just make the human more productive; it fundamentally changes the cost structure of the task. The output is produced for about 10% of the original cost, making the standalone human economically uncompetitive.

Given the massive economic advantages, why wouldn't nations or corporations agree to slow down AI development to manage the transition? The second premise argues this is impossible due to a classic game theory problem: the prisoner's dilemma.

Imagine two competing companies. If both agree to limit their use of AI, they maintain the status quo. But if one company defects and secretly develops AI, it gains an insurmountable competitive advantage. Knowing this, the other company has no choice but to pursue AI as well, just to keep up. Neither can trust the other to hold back.

This dilemma scales globally. Any nation, company, or research lab that pauses AI development risks being left behind by those who don't. The rational choice for every individual actor is to push forward as fast as possible, making collective restraint a near-impossibility.

The final premise addresses the value of human labor. Historically, economic value was tied to the ability to solve problems. In computer science, some problems are considered 'easy' to solve (class 'P'), while others are 'hard' (class 'NP'). For NP problems, finding a solution is difficult, but verifying a proposed solution is easy.

Think of a Sudoku puzzle. Solving a hard one can take a lot of time and brainpower (high NP cost). But if someone gives you a completed grid, you can verify if it's correct in seconds (low P cost).

Modern AI is becoming incredibly adept at solving complex, NP-style problems. It can generate solutions to challenges in science, engineering, and business that were once the domain of highly-paid human experts. As AI's ability to solve these problems approaches the trivial cost of verification, the market value of human problem-solving skills plummets.

This doesn't mean humans become useless. Instead, the locus of value shifts. The critical human skill is no longer finding the answer, but asking the right questions, defining the problem correctly, and verifying that the AI's proposed solution is sound, ethical, and aligned with the intended goals. The role of the human moves from producer to director.

Together, these three premises suggest a future where the relationship between labor, cost, and value is fundamentally reconfigured. The Discontinuity Thesis provides a framework for understanding why the AI revolution may not follow the patterns of the past.