16 Nobel Laureates Say We Have Years, Not Decades, to Adapt to AI

More than two hundred researchers and economists, sixteen of them Nobel laureates, have endorsed "We Must Act Now." The document, organized by Stanford, is an urgent call because it condenses into three ideas what used to be scattered debates across labs, universities, and governments: artificial intelligence will become radically more powerful in less than ten years, the resulting economic shift will be the broadest and fastest ever recorded, and institutional preparation must begin immediately.

What sets this text apart isn't just the number of signatures. It's who's signing. Jeff Dean of Google, Jack Clark of Anthropic, and Noam Brown of OpenAI are the ones building the systems the document warns about. When the people assembling the machines ask for external brakes, the message carries different weight. Anton Korinek, an economist at the University of Virginia, put it plainly: steam, electricity, and computers gave societies decades to adapt; AI might give us only a few years.

That observation gets at the core of the problem. The difficulty ahead comes from squeezing into a few years an institutional adjustment that used to take generations. It's not that this technology is inherently more transformative than the steam engine. The breaking point lies in how much time unions, education systems, and social protection schemes have to reorganize. Institutions, by design, move slowly. That slowness is what makes them institutions.

"The Generosity in the Doorway" already explored this dynamic from the angle of infrastructure that precedes policy. It shows how a small group of actors builds the technical foundation first and only later shapes the response to its side effects. The pattern repeats with names like Sam Altman or consortiums like Stargate. "We Must Act Now" confirms it, though with a twist: some of these same builders are now asking the state to draw the containment lines.

Why does this matter more than it seems? It raises a question the book left open. When the very people accelerating the technology publicly acknowledge its risks, the conversation shifts tone. That doesn't necessarily mean contradiction. It can reflect genuine sincerity alongside incentives that remain fully intact: Jeff Dean can support regulatory frameworks while his organization keeps moving forward without them every quarter. The real doubt isn't whether they're lying, but whether the speed of their creations will outpace the policies they're calling for.

The Marienthal study becomes more relevant here than the Stanford document itself. Marie Jahoda, Paul Lazarsfeld, and Hans Zeisel documented in the 1930s what happened when an Austrian textile factory closed and left an entire town unemployed. The central finding wasn't poverty, which was predictable. It was the collapse of the structure of time: without work, people lost their schedules, stopped reading the newspaper, and even changed the way they walked down the street. Identity fractured before the bank account did.

"We Must Act Now" talks about safety nets and labor policy as if the problem were purely one of income. Marienthal suggests that diagnosis is incomplete. You can hand out a universal basic income and still leave unresolved the question of what a human being does with their days once the structure that used to organize them disappears.

This doesn't mean the document is misguided. It means it falls short if read as a complete solution. A cash transfer addresses survival. It doesn't by itself rebuild the meaning that work used to provide. "The Generosity in the Doorway" insists that the conversation about artificial intelligence and unemployment needs to extend beyond finances, toward rebuilding purpose.

The value of this document lies in the time compression Korinek describes. "Las Piedras No Mienten," a work still in progress, examines how societies that survive abrupt transitions are the ones that designed adjustment protocols without needing broad consensus every single time. The risk of going from decades to just a few years is losing the room to build those protocols gradually. Structures that fail as they grow usually do so because they react under pressure instead of anticipating through design. And shrinking the horizon from decades to years pushes nearly any government into reactive mode.

There's an even more uncomfortable angle. If governments depend on the information that the labs themselves provide about the pace of change, an independent auditing problem emerges. The two hundred-plus signatures and sixteen Nobel laureates lend moral weight, but they don't solve that dependency on the source of information. States need their own measurement capacity, not just corporate press releases.

I'm still not sure whether three years will be enough to build what took the Industrial Revolution three generations. My suspicion is that the timeline is tight. The real value of "We Must Act Now" may not be in answering that question, but in forcing us to ask it out loud, with specific names attached. Stones don't lie, though historians sometimes interpret selectively.

What adjustment mechanisms can we design before this window closes for good?

1. "We Must Act Now" — statement organized by Stanford, signed by more than two hundred researchers and economists, including sixteen Nobel laureates.

2. Statements by Anton Korinek, economist at the University of Virginia, on the comparative timeframes of previous technological revolutions.

3. Jahoda, M., Lazarsfeld, P., Zeisel, H. — Marienthal: The Sociography of an Unemployed Community (1930s study).

4. Yves Laurent, The Generosity in the Doorway (available on Amazon).