Citrini Research: How AI Could Crash the Intelligence Premium and Shake the Economy

Citrini Research has proposed a thought experiment: what happens if, by 2028, AI makes intellectual labor nearly non-scarce? In that scenario, the blow lands not only on the tech market but on the entire middle-class economy, which for decades lived off a premium for knowledge, analysis and the ability to make decisions faster than others.

Where the anxiety comes from

Throughout modern economic history, human intelligence remained a rare asset. Machines could amplify physical labor, speed up production and automate routine, but people were still paid for complex conclusions, strategy, negotiation, design and management. Most of the white-collar economy rested on this: education paid off, experience could be monetized, and mid-level specialists earned a steady markup simply for being competent.

"Human intelligence has always been a scarce resource."

The idea the material points to is that, for the first time, AI hits not muscle or repetitive operations but the very scarcity of thinking. If a model can write, summarize, code, analyze spreadsheets, build presentations and propose solutions, a company can scale mental labor almost like a cloud service. One strong employee with AI can already produce what used to take a team.

How the market is changing

Hence the central nerve: the labor market may begin repricing not because "people are no longer needed" but because the old price of their work becomes too high. Businesses quickly notice the difference between a task that needs unique human judgment and one where a good model operator is enough. When there are many such tasks, pressure falls at once on hiring, pay, career ladders and expectations of office roles.

• Junior and middle positions, with plenty of template-based analytics and material preparation, are squeezed the most.
• Managers lose value if their function boils down to relaying information and monitoring statuses.
• Demand grows for people who can verify model outputs, supply context and take responsibility for the final decision.
• The advantage shifts from "I know how to do it" to "I understand what exactly needs to be done and why."

This does not necessarily mean mass unemployment overnight. Rather, the familiar career ladder starts to break: entry into a profession narrows, while the top is valued even more. If the market used to buy hours of skilled labor, it increasingly buys a result assembled at the junction of people, models, data and automation. That is painful for the middle class, because the predictability of such roles was long the basis of financial stability.

Why a crisis is possible

A global crisis scenario emerges when the repricing hits not one industry but many at once. If millions of knowledge workers begin competing not only with one another but with nearly free AI, employees' bargaining power falls, consumption weakens, and businesses revisit budgets and team structures. At the same time, markets may reprice the future profits of companies that benefit from automation too sharply, and punish those that relied on expensive human labor too quickly.

In this thought experiment, the 2028 date matters more as a signal of speed than as an exact deadline. The worry is not that "everything collapses tomorrow" but that the transition could be abrupt and uneven. Technologies arrive in waves: first they look like toys, then they break the unit economics of entire functions. So it is useful to worry not in the form of panic but in the form of preparation: where do you have unique expertise, where do you own data, where do you have direct access to the customer, and where are you simply selling hours of intelligence that the market will soon learn to buy far more cheaply?

What this means

The main takeaway is simple: AI looks less and less like another useful piece of software and more and more like a factor that reprices human capital. For companies, it is a reason to rethink processes and org structure; for professionals, a reason to move from selling "smart hours" to selling responsibility, taste, domain expertise and the final result.